Field test based guidelines development for the integration of unmanned aerial systems (UASs) in GDOT operations

GEORGIA DOT RESEARCH PROJECT 16-09 FINAL REPORT
Field Test Based Guidelines Development for the Integration of Unmanned Aerial Systems (UASs)
in GDOT Operations
OFFICE OF PERFORMANCE-BASED MANAGEMENT AND RESEARCH
15 KENNEDY DRIVE FOREST PARK GA 30297-2534
April 2019

1.Report No.

2. Government Accession No. 3. Recipient s Catalog No.

FHWA-GA-19-1609

4. Title and Subtitle

5. Report Date

Field-Test-Based Guideline Development for the

April 2019

Integration of Unmanned Aerial Systems (UASs) in 6. Performing Organization Code GDOT Operations

7. Author(s)

8. Performing Organ. Report No.

Javier Irizarry Ph.D. P.E.

16-09

Eric N. Johnson Ph.D.

9. Performing Organization Name and Address

10. Work Unit No.

Georgia Institute of Technology 225 North Ave NW Atlanta GA 30332

11. Contract or Grant No. PI 0016326

12. Sponsoring Agency Name and Address Georgia Department of Transportation Office of Performance-Based Management and Research 15 Kennedy Drive Forest Park GA 30297-2534 15. Supplementary Notes

13. Type of Report and Period Covered Final May 2016 April 2019
14. Sponsoring Agency Code

16. Abstract This project aimed at developing guidelines for the integration of Unmanned Aerial Systems (UASs) in Georgia Department of Transportation (GDOT) operations based on the experience and lessons learned from field tests with UAS technology on selected tasks performed by several groups within GDOT. The research team conducted three focus group sessions with seventeen individuals from the Construction Bridge Maintenance and Intermodal groups. The main purpose of the focus group sessions was to identify tasks that could benefit from the use of UAS and that could be tested in the field. A total of seven locations were selected for field tests including 2 airports 2 rail segments 1 road construction site and 2 bridges. During the field tests several UAS platforms were used to collect various data types including still images infrared images and videos. Flights were performed in both manual and automated modes. After data was collected it was processed and de-briefing meetings were held with participants to collect feedback on usefulness of the process and results. After gaining insights from GDOT personnel who participated in the field tests recommendations for integration guidelines where developed and presented in the accompanying report. The results of this study could lead implementation of UAS from GDOT tasks of the groups included and possibly at GDOT more broadly.

17. Key Words

18. Distribution Statement

Unmanned Aerial Vehicles Unmanned Aerial Systems Operational Requirements Technical Requirements Cost Analysis.

19. Security Classification (of this report)
Unclassified

20. Security Classification (of this page)
Unclassified

21. Number of Pages 22. Price 185

Form DOT 1700.7 (8-69) ii

GDOT Research Project No. 16-09
Final Report
Field Test Based Guidelines Development for the Integration of Unmanned Aerial Systems (UASs) in GDOT Operations
Javier Irizarry Ph.D. P.E. Associate Professor of Building Construction
Eric N. Johnson Ph.D. Professor of Aerospace Engineering
Sungjin Kim Graduate Research Assistant
Kyuman Lee Graduate Research Assistant
Daniel Paes Graduate Research Assistant
Georgia Institute of Technology
Contract with
Georgia Department of Transportation
In cooperation with
U.S. Department of Transportation Federal Highway Administration
April 2019
The contents of this report reflect the views of the authors who are responsible for the facts and the accuracy of the data presented herein. The contents do not necessarily reflect the official views or policies of the Georgia Department of Transportation or of the Federal Highway Administration. This report does not constitute a standard specification or regulation.
iii

Table of Contents
List of Tables.......................................................................................................................................... vii List of Figures ..........................................................................................................................................ix Executive Summary ........................................................................................................................... xiii Acknowledgements .............................................................................................................................. xv List of Acronyms.................................................................................................................................. xvi 1. Introduction.................................................................................................................................1
1.1. Overview .......................................................................................................................................1 1.2. Research Objectives ...................................................................................................................4 1.3. Research Methodology..............................................................................................................4 1.4. Expected Results .........................................................................................................................6 2. Literature Review......................................................................................................................8 2.1. UAS Applications in State Departments of Transportation............................................8 2.2. UAS for Construction Applications.................................................................................... 14 2.3. UAS for Bridge Maintenance Applications ..................................................................... 16 2.4. Image Processing and 3D Models....................................................................................... 18 3. Focus Group Activities .......................................................................................................... 21 3.1. FG Methodology ...................................................................................................................... 21 3.2. FG Results.................................................................................................................................. 26 4. Field Tests of UAS-Assisted Tasks.................................................................................... 46 4.1. Field Test Design ..................................................................................................................... 46 4.2. Field Tests - Aviation Group ................................................................................................ 47 4.3. Field Test Rail Group ......................................................................................................... 69 4.4. Field Test - Bridge Maintenance Group ............................................................................ 71 4.5. Field Test Construction Group ......................................................................................... 75 5. UAS Workshop ......................................................................................................................... 81
iv

5.1 Workshop Attendees ............................................................................................. 81 5.2 Workshop Sessions and Topics ............................................................................ 81 5.3 Results of Group Interview ................................................................................... 85 6. Data Processing ....................................................................................................................... 91 6.1 Photogrammetry Software Selection .................................................................... 91 6.2 Photogrammetry Process with Pix4D ................................................................... 93 7. Data Analysis ..........................................................................................................................107 7.1 Data Analysis Structure and Instruments .......................................................................107 7.2 Performance Factors .............................................................................................................109 7.3 Conceptual UAS-based Workflow ..................................................................... 111 7.4 Usefulness/Suitability Analysis .....................................................................................118 8. Legal Safety and Privacy Considerations....................................................................129 8.1 Title 14 of the Code of Federal Regulation Part 107 ......................................... 131 8.2 Section 333 Exemption Aircraft weighing more than 55 pounds.................... 133 8.3 Certificate of Waiver or Authorization ............................................................... 134 8.4 State UAS Laws .................................................................................................. 135 9. Recommendations for UAS Integration Guidelines.................................................136 9.1 Applicable Regulations Affecting GDOT UAS Integration ............................... 136 9.2 Operational Considerations................................................................................. 137 10. Conclusions and Future Research ..................................................................................145 11. References ...............................................................................................................................150 12. Appendix ..................................................................................................................................158 12.1 Appendix aa IRB Approval Letter..................................................................... 158 12.2 Appendix bb Participant Consent Form............................................................. 159 12.3 Appendix cc Data Collection Sheet ................................................................... 160 12.4 Appendix dd Demographic Information Data Collection Form ........................ 168 12.5 Appendix ee Compilation of Various State UAS Laws .................................... 171
v

12.6 Appendix ff Georgia State UAS Laws .............................................................. 181 12.7 Appendix gg Sample UAS Operations checklists ............................................. 183 12.8 Appendix hh Focus Group Participants ............................................................. 185
vi

List of Tables

Table

Page

1. Table 2-1 UAS Applications Considered by State Departments of Transportation

............................................................................................................................... 13

2. Table 2-2 Summary of Studies on UAS Applications in Construction ............... 16

3. Table 2-3 Recent Studies on UAS Applications in Bridge Maintenance............ 18

4. Table 3-1 FG Sessions with Personnel from Three GDOT Groups .................... 24

5. Table 3-2 Demographic Information of FG Sample............................................ 26

6. Table 3-3 CG Demographic Information............................................................. 28

7. Table 3-4 CG Current Tasks ................................................................................ 29

8. Table 3-5 CG Tasks with UAS Integration Potential .......................................... 30

9. Table 3-6 BMG Demographic Information ......................................................... 32

10. Table 3-7 BMG Current Tasks ............................................................................ 36

11. Table 3-8 BMG Tasks with UAS Integration Potential....................................... 38

12. Table 3-9 IG Demographic Information .............................................................. 40

13. Table 3-10 Current IG Tasks ............................................................................... 43

14. Table 3-11 IG Tasks with UAS Integration Potential.......................................... 44

15. Table 3-12 Potential UAS-assisted Tasks in all Groups...................................... 45

16. Table 4-1 Aviation Group Potential Test Sites .................................................... 48

17. Table 4-2 Site Visit Summary ............................................................................. 49

18. Table 4-3 Selected Test Sites............................................................................... 53

19. Table 4-4 Field Test Code Designation ............................................................... 55

20. Table 4-5 UAS Platforms Specifications............................................................. 59

vii

21. Table 4-6 Roosevelt Memorial Airport Field Test Attendees ............................. 62 22. Table 4-7 Dataset from Field Test ....................................................................... 62 23. Table 4-8 Field Test Attendees............................................................................ 65 24. Table 4-9 Dataset from Field Test ....................................................................... 66 25. Table 4-10 Field Test Designation Codes ........................................................... 72 26. Table 4-11 Field Test Attendees .......................................................................... 72 27. Table 4-12 Dataset from Field Test ..................................................................... 73 28. Table 4-13 Field Test Designation Codes............................................................ 75 29. Table 4-14 Field Test Attendees .......................................................................... 76 30. Table 4-15 Dataset from Field Test ..................................................................... 76 31. Table 5-1 Data Process and Management System Requirements........................ 87 32. Table 5-2 Team Composition Requirements ....................................................... 88 33. Table 5-3 Privacy Safety and Legal Requirements ............................................ 89 34. Table 5-4 Summary of Group Interview Results................................................. 90 35. Table 6-1 Sample of Photogrammetry Software Available................................. 92 36. Table 6-2 Characteristics of Each Processing Options Template........................ 98 37. Table 6-3 Uses of the Results ............................................................................ 106 38. Table 7-1 Demographic Information of Debriefing Session Participants ......... 109 39. Table 7-2 Estimated Workflow Operating Times.............................................. 118 40. Table 8-1 Main Requirements of 14 CFR Part 107 .......................................... 132 41. Table 8-2 Designated Airspaces in United States (Adapted from FAA (2016)) 133
viii

List of Figures

Figure

Page

1. Figure 1-1 Work plan flowchart ............................................................................ 5 2. Figure 3-1 Focus Group Data Collection Process ............................................... 23 3. Figure 3-2 Focus Group Session with GDOT District 1 Construction Division . 27 4. Figure 3-3 Focus Group Session with GDOT Bridge Maintenance Division ..... 31 5. Figure 3-4 Bridge Structure Components ............................................................ 33 6. Figure 3-5 BMG Work Structure......................................................................... 34 7. Figure 3-6 Focus Group Session with GDOT Intermodal Division .................... 39 8. Figure 3-7 Work Organizational Structure (IG) .................................................. 41 9. Figure 4-1 Proposed Field Test Design ............................................................... 47 10. Figure 4-2 Location of the Potential Test Sites ................................................... 49 11. Figure 4-3 Logistics features at Each Airport...................................................... 50 12. Figure 4-4 Images from Visit to Habersham County Airport.............................. 51 13. Figure 4-5 Images from Visit to Monroe-Walton County Airport ...................... 52 14. Figure 4-6 Images from Visit to Newnan-Coweta County Airport ..................... 52 15. Figure 4-7 Field Test Protocol Aviation Group................................................ 57 16. Figure 4-8 UAS platforms used in field tests ...................................................... 58 17. Figure 4-9 Data Collection Plans......................................................................... 60 18. Figure 4-10 Roosevelt Memorial Airport Location and Views ........................... 61 19. Figure 4-11 Field Test Setup ............................................................................... 63 20. Figure 4-12 Field Test Product Samples.............................................................. 64 21. Figure 4-13 Habersham County Airport Location............................................... 65
ix

22. Figure 4-14 Field Test Setup ............................................................................... 66 23. Figure 4-15 Field Test Initial Products ................................................................ 67 24. Figure 4-16 Data collection session at Habersham Airport with GDOT personnel
operating UAS. ..................................................................................................... 68 25. Figure 4-17 Sample products of first rail location test ......................................... 69 26. Figure 4-18 Sample products of second rail location test..................................... 70 27. Figure 4-19 17th Street Bridge Test Location ...................................................... 71 28. Figure 4-20 Image-based Bridge Inspection........................................................ 73 29. Figure 4-21 Sample products of second bridge inspection test site...................... 74 30. Figure 4-22 US129 Project Location ................................................................... 75 31. Figure 4-23 Road Construction Environment...................................................... 77 32. Figure 4-24 Field Test Sample Products.............................................................. 77 33. Figure 4-25 Field Test Sample Products.............................................................. 78 34. Figure 4-26 Field Test Sample Products.............................................................. 79 35. Figure 4-27 Data collection session at US-129 project with GDOT personnel
operating UAS. ..................................................................................................... 80 36. Figure 5-1 Workshop setting ............................................................................... 82 37. Figure 5-2 Brainstorming Session Setting ........................................................... 83 38. Figure 5-3 Hands-on Activity Session................................................................. 84 39. Figure 5-4 UAS platform used for the hands-on activity .................................... 85 40. Figure 6-1 Data Processing Workflow ................................................................ 94 41. Figure 6-2 New Project Window ......................................................................... 95 42. Figure 6-3 Image Properties Window.................................................................. 96
x

43. Figure 6-4 Processing Options Template Window.............................................. 98 44. Figure 6-5 Map View Window.......................................................................... 100 45. Figure 6-6 Processing Bar (Initial Processing) .................................................. 100 46. Figure 6-7 Quality Check .................................................................................. 101 47. Figure 6-8 Preview of Orthomosaic and Corresponding DSM ......................... 102 48. Figure 6-9 Absolute Camera Position and Orientation Uncertainties ............... 103 49. Figure 6-10 Computed Image Positions with Links between Matched Images 103 50. Figure 6-11 Verification of Ground Control Points........................................... 104 51. Figure 6-12 Processing Bar (Point Cloud and Mesh) ........................................ 104 52. Figure 6-13 Processing Bar (DSM Orthomosaic and Index)............................ 105 53. Figure 7-1 Debriefing Sessions Setting ............................................................. 108 54. Figure 7-2 Relevance of Performance Factors .................................................. 111 55. Figure 7-3 UAS-based Workflow vs. Existing Workflow ................................ 112 56. Figure 7-4 Pre-flight Stage ................................................................................ 113 57. Figure 7-5 Flight Stage ...................................................................................... 115 58. Figure 7-6 Usefulness/suitability of Visual Data (CG) ..................................... 119 59. Figure 7-7 Best-suited Viewpoints for Data Collection (CG) ........................... 120 60. Figure 7-8 Usefulness/suitability of 2D Data (IG) ............................................ 121 61. Figure 7-9 Best-suited Viewpoints for Data Collection (IG) ............................ 122 62. Figure 7-10 Usefulness/suitability of 2D Data (BMG) ..................................... 122 63. Figure 7-11 Best-suited Viewpoints for Data Collection (BMG)...................... 123 64. Figure 7-12 Adequacy of Team Composition ................................................... 124 65. Figure 7-13 Adequacy of UAS-based Workflow .............................................. 125
xi

66. Figure 7-14 Efficiency of UAS-based Workflow.............................................. 126 67. Figure 7-15 Safety Improvements (Public and UAS Team).............................. 127 68. Figure 7-16 Summary of Analyses .................................................................... 128 69. Figure 9-1 Examples of mobile applications for airspace reviews before UAS
operations ............................................................................................................ 139
xii

Executive Summary
In April of 2016 a team from the Georgia Institute of Technology entered into a research project to develop guidelines for the use of Unmanned Aerial Systems (UASs) in Georgia Department of Transportation (GDOT) operations. These guidelines would be based on the lessons learned from field tests with personnel from the Intermodal Bridge Maintenance and Construction groups at GDOT. Unmanned Aerial Systems are comprised of a control station for a human operator and one or more Unmanned Aerial Vehicles (UAVs). The utilized UAVs can be equipped with various sensors such as video or still cameras including far and near infrared radar or laser based range finders or specialized communication devices. The ground stations utilized by the human operators can vary from portable computer based systems to fixed installations in vehicles or dedicated control rooms. Several off-the-shelf UAS devices were employed in the research study. These included multirotor as well as fixed wing platforms.
The project lasted for a period of two-years and the research team conducted focus group sessions with seventeen individuals from the three GDOT groups included in the study. The results of these sessions allow the research team to identify the tasks that would be used for field testing with UAS integration. A total of seven locations were selected for field tests including 2 airports 2 rail segments 1 road construction site and 2 bridges. During the field tests several UAS platforms including quadcopters hexacopters and fixed wings were used to collect various data types including still images infrared images and videos. Flights were performed in both manual and automated modes with the use of
xiii

mission planning applications. After data was collected it was processed with photogrammetry software. De-briefing meetings were held with study participants from each of the three groups to collect feedback on usefulness of the process and results. After gaining insights from GDOT personnel who participated in the field tests recommendations for integration guidelines where developed. The recommendations consider the Federal Aviation Administration s regulations as of the writing of the final report. The guidelines address UAS operations planning and execution equipment and data management and UAS operating personnel requirements. The results of this study could complement GDOT s plans for implementation of UAS technologies into operations that can benefit from improved safety of personnel as well as efficient use of resources.
Keywords Unmanned Aerial Vehicles Unmanned Aerial Systems Integration Guidelines Inspections.
xiv

Acknowledgements
The research team would like to acknowledge the Georgia Department of Transportation for its support of the research project and want to thank the following individuals for their direct involvement and contributions to the project
Members of the GDOT Technical Advisory Committee Binh H. Bui Brennan Roney Colette Edmisten
Sungjin Kim Graduate Student Georgia Institute of Technology Kyuman Lee Graduate Student Georgia Institute of Technology Daniel Paes Graduate Student Georgia Institute of Technology Personnel from the following areas
Construction Harold D. Mull Jeana Beaudry Luis Alvarez Intermodal Colette Edmisten Alan Hood Bridge Maintenance Bob O Daniels Joshua Cofer Ryan Beasley Charles
Blue Risk Management and Legal Robert Maguire Annette Simelaro HERO John Sibely Daniel Harneir Industry Partner - Skysight Imaging Rick Dobbins
xv

2D 3D AE ASHT BC BMG CAD CFR CG COA DEM DOT DSM FAA FC FDOT FG FHWA FOI GCP GCS GDOT GIS GPS GT HERO IDOT IG

List of Acronyms
Two Dimensions / Two-dimensional Three Dimensions / Three-dimensional Aerospace Engineering (School) Arkansas State Highway and Transportation Department Building Construction (School) Bridge Maintenance Group Computer Aided Design Code of Federal Regulations Construction Group Certificate of Waiver or Authorization Digital Elevation Model Department of Transportation Digital Surface Model Federal Aviation Administration Facility Coordinator Florida Department of Transportation Focus Group Federal Highway Administration Features of Interest Ground Control Point Ground Control Station Georgia Department of Transportation Geographic Information System Global Positioning System Georgia Tech Highway Emergency Response Operators Illinois Department of Transportation Intermodal Group
xvi

KDOT LiDAR MDOT MnDOT NAS NCDOT NDE NHDOT ODOT PE PIC PMC RABIT TSB TSP UAS UAV UDOT VDOT VO VTOL WSDOT

Kansas Department of Transportation Light Detection and Ranging System Michigan Department of Transportation Minnesota Department of Transportation National Airspace System North Carolina Department of Transportation Non-Destructive Evaluation New Hampshire Department of Transportation Ohio Department of Transportation Project Engineer Pilot In Command Person Manipulate Control Robotic Assisted Bridge Inspection Tool Time Synchronization Board Traveling Salesman Problem Unmanned Aerial System Unmanned Aerial Vehicle Utah Department of Transportation Virginia Department of Transportation Visual Observer Vertical Take-Off and Landing Washington State Department of Transportation

xvii

1.1. Overview

1. Introduction

Unmanned Aerial Systems (UASs) are increasingly being considered for government and civilian applications in the United States. In 2015 the Federal Aviation Administration (FAA) established policies and certification requirements for UAS integration into the National Airspace System (NAS). However a number of issues impede the integration of unmanned aircraft into the manned airspace. Currently unmanned aircraft are allowed to operate under specific conditions that comply with established regulations. Exceptions to these regulations are determined on a case-by-case basis through the FAA waiver framework. Nevertheless the effectiveness of UASs in civil applications has not been clearly determined under these conditions specifically in tasks such as those performed by the Georgia Department of Transportation (GDOT). Aiming at understanding and determining the advantages and limitations of UAS adoption by GDOT as well as its legal societal and operational implications this research conducted various controlled tests (i.e. structured interviews surveys field tests and other activities). The results of these tests were used to develop recommendations for FAA-compatible guidelines for integrating those systems into GDOT operations.

UASs were first widely adopted in military operations and now occupy a permanent position in the military arsenals of many countries (Nisser and Westin 2006). Current civilian applications of such systems include the following
border patrol search and rescue
1

damage assessment during or after natural disasters (e.g. hurricanes earthquakes tsunamis)
locating forest fires identifying farmland frost conditions monitoring criminal activities mining activities advertising scientific surveys securing pipelines and offshore oil platforms (Anand 2007). Several other studies have investigated the application of UASs in the agriculture forestry archeology architecture and construction industries.
A UAS consisting of a rotary wing aircraft with several sensor devices and the ability to hover for extended periods is a well-suited platform for studying UAS applications e.g. autonomous surveillance/navigation (Krajnk et al. 2011) human-machine interaction (Ng and Sharlin 2011) and sport training assistance (Higuchi et al. 2011). In a study conducted by Irizarry et al. (2012) a UAS quadcopter was used to explore the benefits of providing safety managers with still images and real-time video from a range of locations on a construction jobsite. Another study conducted by Rosnell and Honkavaara (2012) showed how virtual point clouds can be generated from image sequences collected by small UASs. A similar study in Finland by Lin et al. (2013) proposed a novel aerial-to-ground remote sensing system for surveying land scenes of interest. The literature review section of this report presents several examples of such studies.
2

Continuous improvements in UAS functionality and performance create opportunities for applied research on integrating this leading-edge technology into various applications. Several departments of transportation (DOTs) across the U.S. have started to explore the use of UASs for various purposes from tracking highway construction projects and performing structure inventories to road maintenance and roadside environmental condition monitoring among many other surveillance traffic management and safety applications. In early 2013 the Georgia DOT (GDOT) started to investigate which of its operations could be optimized with UAS adoption. Four GDOT divisions with the highest potential of benefitting from UAS technology were identified as Construction Engineering Intermodal and Permits & Operation. In order to determine the operational and technical requirements for the use of a UAS by a given division it must identify its operations/tasks and personnel needs to establish a thorough understanding of its goals work environment and internal decision-making processes. Each division s detailed information was processed into a set of requirements that guide the integration of UASs into its operations. As a result five potential UAS configurations were identified (Irizarry and Johnson 2014).
Most tasks within the GDOT divisions studied are governed by an information-sharing process focused on collecting and supplying relevant information to the involved groups. To establish a better understanding of the work dynamics and environment conditions the research team characterized each task by attributes such as location and completion time. In summary the analysis of GDOT tasks provides insight into the operational and technical
3

requirements for integration of UASs into its divisions (Karan et al. 2014 Gheisari et al. 2015).
1.2. Research Objectives
This research project refers to the second phase of the 2013 GDOT study (hereinafter referred to as Phase 2). The objectives of this study are as follows (1) to determine the technological feasibility of utilizing UASs in the operations of GDOT divisions (2) to understand the advantages and limitations of UAS adoption (as well as its legal safety and privacy implications) for tasks identified from the analysis of GDOT divisions (3) to propose FAA-compatible guidelines for integrating such systems into GDOT operations and (4) to hold a workshop for GDOT personnel about the use of UAS technology for the investigated tasks.
1.3. Research Methodology
The research activities involved deep collaboration with GDOT personnel throughout Phase 2. Figure 1-1 presents a flowchart of the research work plan and is followed by a description of the following related activities
Activity 1- Definition of tasks and selection of UAS platforms for field testing
Activity 2 - Performance of field tests Activity 3 - Usability evaluation and development of UAS integration
guidelines Activity 4 - UAS workshop (concurrent with other phases).
4

Activity 1
Definition of Tasks

Activity 2 Field Testing

Activity 3
Usability Eval. Guideline Dev.

Activity 4 Workshop

Figure 1-1 Work plan flowchart
Activity 1 Definition of Tasks and Selection of UAS Platform In this activity the research team performed focus group (FG) sessions with GDOT personnel in the Intermodal Construction and Bridge Maintenance divisions to define the tasks to be performed during the field tests. From the focus group input the research team generated a detailed list of procedures resources and processes followed by GDOT personnel to complete their respective tasks. Next the team analyzed the tasks to determine which UAS platform and related technology is best suited for integration into each one. The last step of this activity involved the design of field tests to be conducted for selected tasks.

Activity 2 Field Testing of Selected Tasks During this activity the research team developed the schedule for the field tests. The estimated test period and number of tests is presented below in the work plan schedule section. Tests were to be conducted according to the field test design developed in the first activity. To understand the impact of UAS use for the selected tasks parts of these field tests involved data collection on task performance and cost analyses. The research team compared the baseline data obtained during Activity 1 to the field test data. The field tests complied with current FAA regulations applicable to the operation of UASs in the national
5

airspace. The results of the tests have been broadly disseminated and are being used by the FAA to develop future regulations on UAS usage by agencies such as GDOT.
Activity 3 UAS Integration Guideline Development and Use Implication Analysis This activity involved the development of guidelines for the integration of UASs into the tasks tested in Activity 2. The guidelines were developed to observe current FAA regulations but the research team also considered regulations that are currently under development by the FAA. In this phase the team analyzed and reported on the legal and societal implications of UAS integration into the tasks examined in the study and performed in general GDOT operations. Data for this analysis were collected through a literature review surveys and interviews with various groups or stakeholders who may have concerns regarding UAS use.
Activity 4 Workshop Development and Delivery In this activity the research team developed and conducted a workshop for GDOT personnel. The topics of the meeting were informed by the outcomes of the field tests. The four-hour workshop was delivered at Georgia Tech facilities at the end of the first year of the research study. Selection of attendees was coordinated with personnel in the Office of Performance-Based Management and Research.
1.4. Expected Results
The expected results of the research project include the following experience in the integration of UASs into GDOT operations 6

knowledge about the safety and legal implications of GDOT UAS use knowledge about UAS performance and cost implications in selected GDOT tasks UAS integration guidelines for selected GDOT tasks increased GDOT personnel understanding of UAS technology in general.
7

2. Literature Review
This section presents a review of recent and relevant UAS initiatives undertaken by DOTs across the United States as well as an overview of specific UAS efforts in key areas related to construction and bridge maintenance.
2.1. UAS Applications in State Departments of Transportation
The Arkansas State Highway and Transportation Department (Frierson 2013) explored different UAS platforms for real-time traffic monitoring and the inspection of highways and bridges. The study did not include field experiments with the platforms due to FAA regulations and schedule constraints.
In a Caltrans study in 2014 the primary goals were to learn more about the use of UASs in geotechnical field investigations and to better understand the legislative issues involved (Karpowicz R. 2014). The report developed a discussion of the role of FAA regulations and a review of other state agencies studies on UAS applications. The study recommends that proof-of-concept testing be conducted in advance of using UASs in transportationrelated tasks and field inspections. In 2008 Caltrans designed a Vertical Takeoff and Landing (VTOL) aerial robot named "Aerobot" for inspection of bridges and elevated structures. At the time the goal was to investigate and improve the robot s capabilities and performance. However due to implementation issues it was never tested in the field (Moller 2008).
The Florida Department of Transportation developed an approach to using UASs for 8

inspections of bridges and high mast luminaires (Otero et al. 2015). The approach involved using a small UAS equipped with high-resolution cameras to provide real-time data. FDOT also conducted proof-of-concept tests to gain insight into the limitations of the proposed approach. The study assessed the UAS platform components and data quality under varying conditions such as altitude payload and maneuverability. Major outcomes included a set of structured UAS-based maintenance procedures as well as estimates of operator training times and of inspection equipment and editing costs.
The Georgia Department of Transportation looked into the economic and operational benefits of using UASs in its operations (Irizarry and Johnson 2014). The study began with the definition of all GDOT division operations that could benefit from UAS use. Semistructured interviews were conducted with potential GDOT UAS operators in order to identify their goals all major decisions involved in accomplishing these goals and the information required in the decision-making process. Several UAS platform configurations were recommended for potential application.
The Michigan Department of Transportation assessed five different UAS platforms comprising optical thermal and LiDAR sensors in various applications (e.g. bridge inspection roadway asset inspection and traffic monitoring). The researchers performed field tests at two bridges two pump stations two traffic monitoring sites and one roadway asset site (Brooks et al. 2015). The department developed an implementation action plan (IAP) encompassing bridge roadway and confined inspections as well as traffic monitoring with improved LiDAR and thermal data processing. The study recommended
9

eight UAS-related topics for future research 1) Operations and maintenance uses and costs 2) Data processing and analysis 3) Slope stability assessment 4) More formal crash scene imaging 5) Aerial imaging to meet MDOT survey supports 6) Optimal methods to store and share large data sets 7) Improvements in thermal imaging and 8) Improvements in UAS positioning.
The Minnesota Department of Transportation also investigated a UAS-based bridge inspection method (Zink and Lovelace 2015). The researchers identified four bridges in Minnesota for field tests with various UAS platforms to evaluate safety issues FAA rules and inspection methods. The three formats of the visual assets collected were as follows 1) still images 2) videos and 3) infrared images. The research also involved the development of 3D models of bridge elements and site locations. The study found that UASs are indeed effective tools for providing critical information for planning costeffective large-scale inspections.
The Ohio Department of Transportation tested UASs for the collection of aerial imagery and developing 3D models of sites (Fred 2013). The 3D point cloud representations of surfaces improved site visualization and analysis. The researchers used the Pix4D software application to process the data into highly geospatially accurate orthorectified images. These images were then added to the ODOT Geographic Information System (GIS) database. Future ODOT plans involve exploring different UAS platforms for bridge condition assessment.
10

A study by the Utah Department of Transportation focused on UAS use on highway projects (Barfuss et al. 2012). The researchers employed a UAS to collect aerial images during and after the completion of a highway corridor project allowing UDOT to develop a visual chronological record of the construction process. The high-resolution images were also used to update the department s GIS database and to identify wetland plant species at Lake Utah. The study concluded that UASs are indeed efficient tools for the collection of real-time data and the documentation of the construction process.
West Virginia University and the Virginia Department of Transportation together developed and tested a UAS named "Foamy " which had been designed for jobsite management and traffic monitoring (Gu 2009). Two field tests of the UAS found a significant number of positioning estimation errors. The researchers performed error analyses to identify the factors affecting positioning accuracy. To improve accuracy a time synchronization board (TSB) was added to the UAS.
The Washington Department of Transportation conducted field tests with UASs on hills above state highways (McCormack and Trepanier 2008). Specifically the department s maintenance division tested a UAS for avalanche monitoring with the aim of preventing accidents and possible highway closures. During the field tests the UAS was able to capture useful aerial images for traffic surveillance.
A study by the North Carolina Department of Transportation explored possible UAS applications on its state highways (NCDOT 2016). The study provided up-to-date
11

information on FAA regulations and developed a guide titled Temporary Flight Restrictions and Aeronautical Charts. The study helped ensure that UAS operators and researchers could understand and comply with UAS-related FAA rules.
The Illinois Department of Transportation developed state regulations for UAS operations (Bryant et al. 2016). In addition to formulating these regulations IDOT also examined UAS applications FAA rules insurance alternatives and safety and privacy issues.
The Kansas Department of Transportation also explored the integration of UASs into their operations (McGuire et al. 2016). Their experience suggests that UASs are useful in bridge inspection radio tower inspection surveying road mapping high mast light tower inspection and stockpile measurement. The study also conducted a survey and an analysis of strengths weaknesses opportunities and threats to determine how to improve the safety and efficiency of UAS operations.
The New Hampshire Department of Transportation looked into how to increase safety and efficiency while reducing the operational costs of UASs (Hunt 2016). The study focused on integrating UASs into monitoring traffic and assessing infrastructure conditions to improve these tasks. It also aimed to educate NHDOT employees on how to use the technology. Table 2-1 lists these UAS-related studies and others conducted by state departments of transportation.
12

Table 2-1 UAS Applications Considered by State Departments of Transportation

DOT Arkansas California Florida Georgia Michigan Minnesota
Ohio
Utah
Virginia
Washington North Carolina
California Illinois Kansas
New Hampshire

Applications
Real-time traffic movement monitoring and highway bridge and facilities inspection
Geotechnical field investigations Bridge and high mast luminaires (HMLs) inspection

References
Frierson 2013
Moller 2008 Otero Gagliardo Dalli Huang and Cosentino 2015

Monitoring remote and rural areas in Florida Economical and operational benefits of UAS integration into DOT operations
Bridge inspection traffic monitoring or roadway asset surveillance
Bridge inspection Three-dimensional model based on visual data collected with a UAS and Geographical Information System for project planning

Werner 2003 Irizarry and Johnson 2014 Brooks Dobson Banach Dean Oommen Wolf Havens Ahlborn and Hart 2015 Zink and Lovelace 2015
Fred 2013

Data collection on freeway conditions intersection movement network paths and parking lot monitoring Taking high-resolution pictures of highways to inventory their features and conditions quickly and at a very low cost Transportation worksite inspection and traffic monitoring

Coifman et al. 2004
Barfuss Jensen and Clemens 2012 Gu 2009

Real-time traffic surveillance monitoring of traffic incidents and signals and assessment of environmental conditions of roadside areas Highway maintenance and traffic surveillance
Capturing aerial images for data collection and traffic surveillance on mountain slopes above state highways
UAS operator guidelines
To develop a vertical takeoff and landing (VTOL) aerial robot called an Aerobot for elevated structure inspection To understand the UAS concept regulatory and operational issues and the safety and privacy concerns of implementation
To develop and provide recommendations for safer and more efficient UAS use on DOT tasks
To analyze the cost benefits and human factors of UAS integration on DOT tasks

Carroll and Rathbone 2002 McCormack and Trepanier 2008
Coifman et al. 2004 North Carolina Department of Transportation 2016 Moller 2008
Bryant et al. 2016
McGuire et al. 2016
Hunt 2016

13

2.2. UAS for Construction Applications
Unmanned aerial systems are increasingly being considered for applications in the construction environment. This section reviews research of such applications.
Hart and Gharaibeh (2011) conducted field tests with UASs on ten roadways in Texas to determine whether UAS use would improve the safety of roadside conditions and the accuracy of construction inventory surveys. Roadside conditions were evaluated by the examination of visual data collected with the UAS. Weather and field conditions were identified as major variables affecting overall UAS performance.
Blinn and Issa (2016) explored possible applications of UASs in active construction environments. Their study compared traditional task performance (without UASs) to UASsupported task performance. They found that visual data provided by the UAS is indeed useful in project management and control on construction sites. In addition the study showed that the use of a UAS for certain tasks was superior to traditional methods since it could decrease operational costs.
Irizarry and Costa (2016) also investigated possible UAS uses in construction management. The study involved collecting qualitative and quantitative data through interviews with and surveys of construction managers. The findings indicate that construction progress monitoring and jobsite logistics could benefit from the visual assets captured and provided by the UAS.
14

Kim et al. (2016) identified performance factors user requirements and operational challenges associated with the use of UASs for construction site inspections--particularly for safety inspections on jobsites. A survey questionnaire was distributed to safety and project managers in the field. A total of 31 factors and 17 measures were identified and used to evaluate the performance of UAS operations. Flight plans and documentation methods were determined to be the most critical user requirements whereas FAA regulations and pilot certification were considered the most significant challenges for safe UAS operations in construction environments.
Gheisari and Esmaeili (2016) identified user and technical requirements for UAS safety applications. Safety managers indicated the following hazardous operations as the ones that would benefit the most from UAS use 1) working around traffic or cranes 2) working near an open area and 3) working in the blind spot of heavy equipment. The three most critical technical requirements identified were as follows 1) real-time communication 2) a highprecision navigation system and 3) a sense-and-avoid system. Table 2-2 below summarizes the studies on UAS applications in construction.
15

Table 2-2 Summary of Studies on UAS Applications in Construction

References
Blinn and Issa 2016
Gheisari and Esmaeili 2016
Kim et al. 2016
Irizarry and Costa 2016 Rinaudo Chiabrando Lingua & Span 2012 Eisenbei & Zrich 2009 Hudzietz & Saripalli 2011 Barazzetti Remondino & Scaioni 2010 Hart and Gharaibeh 2011
Metni & Hamel 2007 Irizarry et al. 2012 Eschmann Kuo Kuo & Boller 2012

Objectives To provide the potential uses of UASs on construction environments through a survey of construction industry professionals To identify user and technical requirements for using UASs for safety management tasks To identify user requirements operational challenges and performance factors of UAS use in construction To identify potential applications of visual assets obtained from a UAS for construction management tasks
To monitor daily activity of excavation work
To collect terrestrial images
To create 3D models of trains
To create 3D models of structures
To evaluate the effectiveness of UAS to collect condition data of roadside infrastructure. To inspect bridges
To improve safety management
To detecting cracks in buildings

2.3. UAS for Bridge Maintenance Applications
Bridge maintenance activities are considered an ideal UAS application. This section reviews several research efforts to evaluate this application. Menti and Hamel (2007) studied the adoption of a UAS equipped with a computer vision sensor for bridge monitoring. This UAS deployed a novel UAS control method for quasi-stationary flights above each bridge monitored. Guerrero and Bestaoui (2013) employed the ZermeloTraveling Salesman Problem (TSP) method to generate optimal flight routes for bridge structure inspections. The TSP method was able to improve overall flight performance depending on weather conditions. Hallermann and Morgenthal (2014) explored autonomous and semi-autonomous flights for detecting structural damage on bridge
16

structures. La et al. (2014) employed a robotic system for autonomous bridge deck inspections. The navigation system was designed to collect and conduct a non-destructive evaluation (NDE) of visual assets. The system aimed to reduce costs time and risks associated with bridge deck inspections. Khan et al. (2015) tested a UAS for inspecting bridge structures in inaccessible locations. Tests were conducted initially on a mock-up bridge model and then on real highway bridges. The researchers found that future research could involve the development of computer vision-based UASs.
Chan et al. (2015) reviewed the current state of UAS-based bridge inspections. The study looked into the technology s historical development inspection performance and requirements. They conducted a case study to analyze the cost effectiveness of UAS-based inspections and found that around US$3 000 of inspection costs could be saved from reduced traffic control and resources in general on a construction project. Gucunski et al. (2015) designed and validated the performance of the Robotic Assisted Bridge Inspection Tool (RABIT). Gillins et al. (2016) designed a protocol for UAS-based bridge inspections as a result of field tests on a bridge in Oregon. Table 2-3 summarizes these recent studies on UAS applications in bridge maintenance.
17

Table 2-3 Recent Studies on UAS Applications in Bridge Maintenance

References Hallermann and Morgenthal 2014 Laa et al. 2014
Metni and Hamel 2007
Chan et al. 2015
Gucunski et al. 2015 Gillins et al. 2016 Guerrero and Bestaoui 2013
Khan et al. 2015

Objectives To develop a method of visual bridge inspection based on aerial photos and video taken by a UAS To explore how visual data collected with a UAS can be used to inspect the bridge deck conditions of common highway bridges
To study the UAS application for monitoring bridge maintenance with a computer vision sensor
To provide an overview of UAS-based visual bridge inspection studies and to address the obstacles to integrating this technology into current practice To develop and implement the RABIT system (Robotics Assisted Bridge Inspection Tool) To develop a methodology for bridge inspection in Oregon using a UAS To develop a methodology for developing structure inspection-based simulations To evaluate bridge conditions with a UAS equipped with a set of remote sensors to conduct a mock-up test on a small concrete bridge model and an actual small/medium bridge

2.4. Image Processing and 3D Models
UAS images can be processed to create three-dimensional models of the objects or areas captured. This section reviews research showing how these 3D images are among the most useful products of visual data collected by UASs. Oskouie et al. (2015) developed a framework that integrated images and point cloud processing to produce high quality data for project condition assessment. A field test to validate the proposed framework deployed an off-the-shelf UAS platform to collect aerial images of an academic building at the University of Southern California. Commercial photogrammetry software applications were used to create a 3D model of the building. During data processing geometric features of interest (FOIs) were detected extracted and localized within the 3D point cloud to improve the accuracy of the features classification. The researchers concluded that to validate the framework further field testing with more detailed parameters is required.

18

Siebert and Teizer (2014) describe the main UAS components needed to generate flight plans including hardware and software applications. Their study demonstrates how such technology can use aerial data collection and processing to generate 3D point clouds orthomosaic maps and digital elevation models. The research team conducted three case studies to assess the efficiency of UAS-based construction earthwork surveys on three distinct types of jobsites landfill road construction and rail construction. The UAS performance was evaluated on the basis of the error analysis results and then compared to the conventional survey method. Findings indicated that the 3D point clouds generated from the UAS data enabled more accurate measurements of the earthworks compared to the traditional earthwork survey process. The study also identified some technical limitations of UAS-based surveys e.g. the UAS battery life and camera/image resolution. The research team recommended case studies of various scenarios in order to identify other possible UAS applications in construction work environments.
Rodriguez-Gonzalvez et al. (2014) proposed a methodology to reconstruct 3D models from aerial images obtained during UAS flights. The methodology involves computer vision processing and photogrammetric algorithms to extract and match key points from multiple images. The 3D model is then reconstructed through image orientation. Field tests were performed to validate this UAS-based method and assess the quality of the reconstructed 3D models. The research team found that the method was more cost effective and provided more accurate 3D models compared to those generated from terrestrial laser scanners.
19

d Oleire-Oltmanns et al. (2012) explored UAS use for more accurate and reliable 3D data for soil erosion control. A fixed-wing UAS with satellite-based remote sensors were used to collect data on different test environments. Installed ground control points (GCPs) allowed for the geo-referencing and processing of the collected images to generate the 3D models. The GCP-based workflow allowed for the development of highly accurate 3D models and the approach proved to be very efficient for erosion assessments.
Ellenberg et al. (2014) conducted a feasibility study on potential UAS applications for infrastructure inspections. To enable a UAS to identify markers placed on an inspected structure the research team developed an image-processing algorithm that provided distances and angles between the aerial vehicle and the markers. During the study the researchers performed two lab tests and one field experiment to evaluate how well the system detected defects and damage on a bridge structure. The developed system was proven to more effectively generate accurate data than the traditional human-based inspection method. The study contributed to a better understanding of how UASs and image-processing algorithms can be combined and integrated into infrastructure inspection. This review of the research aims at providing context for the reader on the topic of UASbased data. This larger view can help GDOT personnel assess the selected applications considered in this research project.
20

3. Focus Group Activities
This chapter presents data obtained from three focus group (FG) sessions with GDOT personnel (one session per each participating GDOT division). The sessions took place between mid-July and early August of 2016 and lasted between two and three hours each. At the beginning of each session the general research goals and the objectives of the session were explained to all participants. The attendance sign-up sheet (with individual code numbers for identification) the demographics form and the data collection sheet were distributed to all participants. A total of 17 management-level professionals participated in the FG sessions distributed as follows five from Construction seven from Bridge Maintenance and five from Intermodal.
3.1. FG Methodology
A FG session is a type of group interview that has proven to be an effective method for collecting qualitative data on a specific topic (Kitzinger 1995 Sim 1998). This method is widely used for exploring and examining the nature of participants knowledge and experience providing insight into how participants view a topic or process (Kitzinger 1995) as well as how they might changes their views and what information they might require (Denning & Verschelden 1993).
3.1.1. FG Objectives
In this study the FG sessions provided the research team with information on the participating GDOT divisions current tasks resources and decision-making processes.
21

This information was found to be critical to identifying the tasks that would benefit from UASs integration (Irizarry et al. 2017). The main research goals involved the following four objectives
1) Compile a list of the participating GDOT divisions current tasks including detailed descriptions of their organizational structures work processes and required resources.
2) Define tasks that can benefit from the use of UAS. 3) Identify general UAS integration requirements such as operational concepts
technological requirements work environment conditions and user characteristics. 4) Develop a field-testing protocol.
3.1.2. FG Data Collection Plan
The data collection plan was designed to achieve the objectives listed above. The FG sessions involved both unstructured and structured interviews. Before conducting the focus group the researchers submitted a consent form (comprising the interview questions procedures benefits and compensation) to the Georgia Tech Institutional Review Board (IRB) for evaluation and approval. The IRB is responsible for ensuring the physical and mental wellness of human research subjects. (See Appendix aa for the IRB approval form Appendix bb for the consent form and Appendix cc for the structured interview questions and Appendix dd for the attendance sign-up sheet and the demographics questionnaire.) Figure 3-1 illustrates the FG data collection plan.
22

Figure 3-1 Focus Group Data Collection Process
3.1.3. FG Participants
In general an FG session includes the coordinators (i.e. a moderator and a facilitator) and the interviewees.
FG Coordinators In this study the FG moderator was in charge of leading the discussions during the FG sessions promoting participants interest in the topic and encouraging them to engage in discussion (Kitzinger 1995 Sim 1998). The moderator was also in charge of introducing the participants to the research objectives FG session goals and general data collection procedures (e.g. the types of information that would be collected). The facilitator was in
23

charge of recording all conversations and making annotations throughout the data collection process to allow for accurate verbatim analysis during the data analysis phase.

FG Interviewees The seventeen GDOT employees (see Appendix hh for list of participants) who had volunteered to participate in the study came from the Construction Bridge Maintenance and Intermodal groups. The FG sessions were conducted separately for each group between mid-July and early August of 2016 and lasted between two and three hours each.

The first FG session with the five participants from the GDOT District 1 Construction Group (CG) took place at their office in Gainesville Georgia on July 12 2016. The second FG session with the seven participants from the Bridge Maintenance Group (BMG) took place at the Georgia Transportation Management Center on July 19 2016. The last FG session with the five participants from the Intermodal Group (IG)--three having come from the Aviation team and two from the Railway team--took place at the GDOT office at the One Georgia Center in Atlanta on August 1 2016. Table 3-1 below summarizes this information on these GDOT FG sessions.

Table 3-1 FG Sessions with Personnel from Three GDOT Groups

FG Participants
Construction Group (CG )
Bridge Maintenance Group (BMG )
Intermodal Group (IG )

Number of Participants
5
7
5

Location
GDOT District 1 Gainesville Office
Georgia Transportation Management Center Atlanta
One Georgia Center Atlanta

Date July 12 2016 July 19 2016 August 1 2016

24

The next sections describe and elaborate on collected data including demographic information and interview outcomes. Data collected from the FG sessions is considered and treated as qualitative data since in this study the groups may not provide the required degree of representativeness (Sim 1998). A total of 17 participants were recruited for the three FG sessions. The FG participants included 14 males (82.4 percent) and three females (17.6 percent) all of whom had worked in infrastructure and construction-related fields for fewer than 10 years (35.3 percent) between 11 and 20 years (29.4 percent) or over 21 years (35.3 percent). The participants ages varied from under 30 years of age (5.9 percent) to over 50 years of age (29.4 percent). Eight participants had high-school diplomas (47.0 percent) seven participants had bachelor s degrees (41.2 percent) and two held master s degrees (11.8 percent).
All participants (100 percent) were familiar with the basic concept of UAS and the idea of integrating this technology into their tasks. However most of them do not have any UAS flight experience. Only three out of 17 participants (17.6 percent) had UAS flight experience for either recreational or research purposes. Two participants from the CG had engaged in UAS flying for recreational purposes and one person from the IG had used an UAS in urban and city planning research. Table 3-2 displays the demographic information of all FG participants.
25

Table 3-2 Demographic Information of FG Sample

Attribute Gender Male Female Age Under 30 years 31-40 years 41-50 years Over 51 years Work experience Less than 10 years 11-20 years Over 21 years Educational Attainment High-school level Undergraduate level Graduate level UAS Knowledge Know Do not know UAS Flight Experience Yes No

Participants (N 17)
82.4% 17.6%
5.9% 41.1% 23.6% 29.4%
35.3% 29.4% 35.3%
47.0% 41.2% 11.8%
100% 0.0%
23.5% 76.5%

3.2. FG Results
This section presents the results of the interviews conducted during the FG sessions with each of the three groups (i.e. the Construction Bridge Maintenance and Intermodal groups). The results include the demographic information of each group its tasks team structure and associated resources as well as the identification of the tasks that could benefit from UAS integration.

26

3.2.1. Construction Group (CG)
CG Demographic Information A total of five individuals (N 5) from the CG participated in the FG session. The group included four males (80 percent) and one female (20 percent) two of them were over 50 years of age (40 percent) and the others were between 41 and 50 (60 percent) years of age. Figure 3-2 presents a photograph of the setting of the FG interview with the CG participants. All participants were responsible for managing road construction projects within the GDOT District 1 Office as project managers (20 percent) or project engineers (80 percent). Moreover they all had significant experience in their current positions or in related fields. Three participants (60 percent) had more than 21 years of experience in their current group. Four participants had more than 21 years of experience (80 percent) in construction-related fields. In regard to educational attainment four participants had high-school diplomas (80 percent) and one had a bachelor s degree in civil engineering. Table 3-3 summarizes the demographic information of the CG participants.

(a) FG Session Introduction

(b) Data Collection from FG Participants

Figure 3-2 Focus Group Session with GDOT District 1 Construction Division

27

Table 3-3 CG Demographic Information

FG Member ID Gender Age Job Position Job Description Experience in Current Position Experience in Related Field Size of Department ( of Employees) Educational Background Education Attainment UAS Knowledge UAS Experience
If yes how long
If yes for what use

C01

C02

C03

C04

Male

Male

Female

Male

Over 50

41-50

Over 50

41-50

Project

Project

Project

Project

Engineer

Engineer

Engineer

Manager

Management of GDOT Road Construction Projects

Over 21

Over 21

Less Than 10 Less Than 10

Years

Years

Years

Years

Over 21

Over 21

Over 21

1120

Years

Years

Years

Years

Very Large Small (Less Large (50 Large (50

(over 100) than 25)

100)

100)

No Major

No Major

No Major

Civil Engineering

High-School Diploma

High-School Diploma

High-School Diploma

Bachelor

Yes

No

No

Yes

Yes

Less than 1 Less than 1

year

year

Recreational Recreational

C05 Male 41-50 Project Engineer
1120 Years Over 21 Years Small (Less than 25)
No Major
High-School Diploma
No

CG Current Tasks The interviewees from the CG all agreed that the main responsibility of a project engineer (PE) is to conduct field surveys take linear and area measurements and verify that contractually required items and construction materials are present at the construction jobsite. Usually PEs will collect videos and photos of the jobsite to facilitate their assessment of the work environment. However such procedures may pose risks to them. For instance when they inspect underground pipelines and ground utilities from the roadside they risk being struck by passing vehicles. Under such conditions the inspector s safety is protected by safety signs placed on the road.

One of the PE s main tasks is to measure concrete and earthwork. The PE is in charge of verifying the volume of earth excavated when the GDOT Construction division processes
28

payments to earthwork contractors. To quantify excavation volume Construction division personnel usually use a simple calculation method involving the multiplication of the height by the square footage of the void in the ground or the multiplication of the number of dump trucks used to remove the soil by their maximum load capacity.

Besides ensuring proper execution of excavations PEs are also responsible for erosion control overseeing project limits and work areas. PEs are required to wear special boots and walk around the excavation area. Using measuring devices to assess erosion. The FG participants considered this a task of special concern. They noted other responsibilities of PEs including the inspection of pedestrian sidewalks and monitoring of traffic speed and flow to prevent hazardous situations and accidents at the jobsite. Table 3-4 summarizes the identified CG tasks.

Group CG

Table 3-4 CG Current Tasks

Tasks

1.

Site monitoring

2.

Volume measurement (earthwork)

3.

Erosion control

4.

Traffic and heavy equipment control

5.

Pipeline and sidewalk inspection (logistic)

CG Tasks with Potential for UAS Integration This section discusses the CG operations that could integrate UASs. In general the integration of UASs into CG operations could lead to major improvements in construction monitoring and documentation especially with respect to frequency data accuracy and safety of CG personnel among others. Earthwork measurements and erosion control

29

inspections are identified as CG tasks that could benefit from UAS adoption (Irizarry et al. 2017). The geo-referenced visual data captured by UASs allows for the development of 3D models through photogrammetry software applications. PEs could rely on these 3D models to quantify excavation volumes and to measure the elevation of work areas for erosion control. Due to its real-time video feed feature a UAS could also assist in traffic control and heavy equipment displacement at/around the construction jobsite. Table 3-5 summarizes the identified CG tasks that could integrate UASs.

Group CG

Table 3-5 CG Tasks with UAS Integration Potential
Tasks 1. UAS-based 3D model
Erosion control Excavation measurement (quantification) 2. High-frequency site monitoring/inspection (daily or weekly inspection) 3. Traffic control and heavy equipment displacement

3.2.2. Bridge Maintenance Group (BMG)
BMG Demographic Information Seven individuals (N 7) from the Bridge Maintenance Group (BMG) attended an FG session. All participants were male. Three of the BMG participants were 50 years of age (43 percent) and four participants were between 31 and 40 years of age (57 percent). One manager (14.3 percent) was in charge of the division two bridge inspection supervisors (28.6 percent) were responsible for monitoring all bridge inspection jobs three bridge inspection specialists (42.8 percent) performed bridge inspections and one bridge inspection technician (14.3 percent) assisted the bridge inspection supervisors in the inspection and decision-making processes. Most respondents have less than 10 years of

30

experience in their positions (85.7 percent) but one participant had between 11-20 years of experience (14.3 percent). In regard to experience in bridge maintenance two participants had over 21 years of experience (28.6 percent) two had between 11-20 years of experience (28.6 percent) and three had less than 10 years of experience (42.8 percent). With respect to educational attainment four participants had high-school diplomas (57.1 percent) three had bachelor s degrees (42.9 percent) two of which were in civil engineering. Table 3-6 summarizes the demographic information of the BMG interviewees. Figure 3-3 illustrates the FG session setting.

(a) FG Session Introduction

(b) Data Collection from FG Participants

Figure 3-3 Focus Group Session with GDOT Bridge Maintenance Division

31

Table 3-6 BMG Demographic Information 32

BM01

BM02

Gender

Male

Male

Age

Over 50

31-40

Job Position State Manager Technician

Job Description

Manage the State Assist

BMG

Supervisor

Experience in Current Position

Less

than

10

years

Less than years

10

Experience in Field

Related

21-25

years

Size of Department 50-100

Less than 10 years
50-100

Educational Background

Biology

Education

B.S.

Attainment

UAS Knowledge Yes

No Major
High-School Diploma

UAS Experience No

BM03 Male 31-40 Supervisor Supervise Inspections
11-20 years
11-20 years
Less than 25
No Major
High-School Diploma

BM04 Male Over 50 Specialist Perform Inspection Less than 10 years
21-25 years
50-100 Civil Engineering
B.S.

BM05 Male
31-40
Specialist Perform Inspection Less than 10 years
Less than 10 years
50-100
Civil Engineering
B.S.

BM06 Male Over 50 Specialist Perform Inspection Less than 10 years
Less than 10 years 50-100
No Major
High-School Diploma

BM07 Male 31-40 Supervisor Supervise Inspections Less than 10 years
11-20 years
Less than 25
No Major
High-School Diploma

BMG Work Environment The main operations of the BMG division involve performing inspections on approximately 15 000 bridges in Georgia. The division consists of three teams with different inspection roles depending on the bridge component to be inspected. Basically a bridge has three main components 1) deck 2) superstructure and 3) substructure (includes areas of bridge located underwater). (See Figure 3-4.)
Figure 3-4 Bridge Structure Components
1. Deck Supports the roadway and traffic also distributes "live" and "dead" loads. 2. Superstructure Supports loads transmitted through the deck. 3. Bearings Support the transfer loads between the superstructure and the
substructure. 4. Substructure Transfers all loads from the superstructure to the ground. 5. Expansion Joints Absorb expansion and contraction of the superstructure and
protect the bearings from water and debris. 33

The BMG develops and adopts internal references and standard protocols for conducting its operations. An example of these materials is the Bridge Structure Maintenance and Rehabilitation Repair Manual (GDOT 2012) which is based on the American Association of State Highway and Transportation Officials guide for bridge inspection and maintenance (AASHTO 2010). Figure 3-5 illustrates the work structure of the BMG.
Figure 3-5 BMG Work Structure BMG Current Tasks The BMG performs visual observations as part of its work to inspect the various elements of a bridge. Usually these visual inspections involve viewing bridge elements from different distances and viewpoints. Depending on the type of bridge structural elements size and traffic on the bridge the inspection task may have a different sequence and frequency. The topside teams conduct regular inspections in two-year cycles for each bridge. (See Figure 3-5.) To accommodate the variability of bridge size location and
34

condition the specialized teams operate on three- six- and 48-month cycles. The underwater teams work on a 60-month cycle to inspect underwater elements.
The GDOT BMG measures vertical clearances and surveys permanent capacity as scheduled. In addition the group uses hammers to inspect connection points in hard-toreach locations. Sometimes an infrared camera is used to detect temperature differences to identify problems with concrete delamination in the deck or caps. A temperature profile can also be used to detect cracks on bridge elements. To ensure the safety of its personnel the group has designed contingency plans for any unforeseen danger or accident during inspections.
To allow for an efficient and safe inspection process the BMG coordinates with authorities with jurisdiction over the bridge (or with third parties such as traffic control companies or Federal Highway Administration (FHWA) units in charge of a road s traffic management) to control traffic flow. A typical inspection process involves the bridge deck team and/or the specialized team. The average time required to inspect a bridge ranges from 15 minutes to three or four hours depending on the structure size and type of bridge. It usually takes the team 15 to 20 minutes to set up the equipment upon arrival at the site. Table 3-7 lists the current tasks of the BMG.
35

Group BMG

Table 3-7 BMG Current Tasks
Tasks
1. Visual observation (sequence and frequency) Depends on bridge type structural system size and road traffic conditions Regular inspection (two years) specialized team (three six or 48 months) underwater team (60 months)
2. Vertical clearance measurement 3. Hammer used to inspect hard to access locations 4. Accident or contingency plan (procedures)
Reports problem to BMG Starts traffic control (takes about 30 minutes) Starts to set up equipment (15-20 minutes) Inspect the point of interest (ranging from 15 minutes to over 4 hours)

BMG Tasks with Potential for UAS Integration The integration of UASs into bridge maintenance operations could save a significant amount of time particularly on inspections of bridges with tall columns. Moreover a UAS is capable of flying underneath bridge decks facilitating the inspection of hard-to-reach structural elements such as bearings connections and column caps. However because satellite signals may be weakened under bridge structures it is likely that the Global Positioning System (GPS) sensors of the UAS will struggle to find and lock on to those signals. Such a technical limitation restricts the use of UASs in bridge inspections. UASs should also be equipped with special cameras with built-in flashlights capable of pointing up and illuminating hard-to-capture elements e.g. the undersides of bridge structures. UASs that are able to generate 3D models--as described above--are ideal for assessing cracks and vertical clearances since their use requires no interruptions of traffic on the bridge roadway. The 3D models also enable bridge maintenance teams to check the accuracy of the original bridge plans and address any significant deviations from the

36

original design. To ensure the safety of any UAS inspection operation the GDOT BMG should check for and avoid power lines around the inspection area.
Another task that could benefit from UAS integration is the inspection of the interior of box-beams conducted to detect cracks on the inside walls. Since the interior of box-beams lack sufficient light for direct observation and the detection of cracks is a visual process performing UAS inspections in such confined spaces requires more caution and time than is required for other bridge elements. Similar to UASs for inspections underneath structures this application would also require powerful cameras with built-in lights besides requiring the capability for more precise manual maneuvers.
Moreover the incorporation of sonar sensors on the unmanned vehicle would enable its use in underwater inspections. Sonar sensors can detect the vehicle s vertical position when it comes into contact with the water surface as well as measure its distance from the bridge deck or from the bottom of the body of water (e.g. river or lake bed). This application could assist divers when performing inspections of submerged elements by checking for debris and other possible entangling hazards. Further analysis would be needed to determine the impacts of using this technology on the time needed to conduct underwater measurements. Table 3-8 summarizes BMG operations that could incorporate UASs.
37

Group BMG

Table 3-8 BMG Tasks with UAS Integration Potential
Tasks 1. Time-saving on bridges with tall columns (an upward-looking camera and strong
light is required) 2. UAS-based 3D model
Crack detection and assessment Vertical clearance assessment 3D steel beam model development for comparison of as-built to original
designs. 3. Inspection underneath bridges and decks using various sensors (e.g. infrared camera or thermal sensor)

3.2.3. Intermodal Group (IG)
IG Demographic Information Participants from four different departments comprise the Intermodal Group aviation railway freight transport system and public transit. However for the FG session only the first two departments were selected (aviation and railway) since the freight transport system and public transit departments are more involved with transportation than with construction matters. Figure 3-6 illustrates the setting of the FG session with the IG. A total of five participants volunteered for this FG session (N 5). Three were from the aviation department (60 percent) and two from the railway department (40 percent) three were male (60 percent) and two were female (40 percent) three were between 31 and 40 years of age (60 percent). Each participant had a different professional role (job position) and had less than 10 years of experience in the current position. However two participants had 11-20 years of experience (40 percent) in airport inspection. In regard to educational attainment three participants (60 percent) had bachelor s degrees in civil engineering or aviation management and two (40 percent) held master s degrees in urban planning or aviation and safety management. Table 3-9 shows the demographic information of the IG.
38

(a) FG Session Introduction

(b) Data Collection from FG Participants

Figure 3-6 Focus Group Session with GDOT Intermodal Division

39

Table 3-9 IG Demographic Information

Participant ID Gender Age
Job Position

I01 Male 41-50
Railway Engineer

Job Description

Railway Inspection

Experience in Current Position
Experience in Related Field
Size of Department
Educational Background
Education Attainment UAS Knowledge UAS Experience If yes how long If yes for what use

Less than 10 years
Less than 10 years Small (fewer than 25)
Civil Engineering
Bachelor
Yes No -

I02 Female 31-40
Railway Planner
Railway Planning
Less than 10 years Less than 10 years Medium (25-50)
Urban Planning
Master
Yes 1-2 years Research

I03 Male Less than 30
Airport Project Engineer
Airport Construction Management
Less than 10 years
Less than 10 years Small (fewer than 25)
Aviation Management
Bachelor

I04 Female 31-40 Airport Program Manager Airport Department Management
Less than 10 years
11-20 years
Small (fewer than 25) Aviation and Safety Management
Master

No

No

-

-

-

-

I05 Male 31-40 Airport Inspection Manager
Airport Inspection
Less than 10 years
11-20 years
Small (fewer than 25)
Aviation Management
Bachelor
No -

IG Work Structure Figure 3-7 presents the organizational structure of the four departments of the GDOT Intermodal Group 1) aviation 2) railway 3) freight transport and 4) public transit. As mentioned previously only the aviation and railway departments participated in the FG sessions.

40

Figure 3-7 Work Organizational Structure (IG)
The railway department has contracts with six consultants that conduct daily inspections on the railways and their surroundings. The aviation department performs its own inspections dividing itself into two subgroups 1) the airport planning team and 2) the airport development team. The airport planning team is mainly in charge of inspecting the pavement conditions of airport runways as well as conducting visual assessments of the general conditions around the airport and its runways. Planning also conducts reviews of airport master plans and layout plans. The aviation department sometimes hires third-party inspectors. The aviation department s other subgroup the airport development team is charged with monitoring the progress of construction at airport facilities. Seven project managers in this group are assigned to the 104 airports in Georgia.
41

Current Tasks of IG The general railway inspection process involves four steps 1) walk through the railway line 2) check its general conditions 3) take pictures of points of interest and 4) document and address issues and document solutions. The railway department has been using a truck equipped with a camera to facilitate such tasks. The truck is used to record videos of the rails and its surroundings at an average speed of five miles per hour 30 to 50 miles per day.
Inspectors and managers in the aviation department drive onto the airport runways and taxiways to perform visual inspections. This task usually requires the use of special equipment such as a range finder an inclinometer and a measuring wheel. The time required to inspect a runway is a function of the size of the airport. Runway inspections involve verifying that the runway markings are visible and signs are intact the height of trees around and in the airport environment meets FAA standards and the conditions of the airfield pavements are maintained. To ensure the safe operation of aircraft obstacles and cracks on the runway should be carefully inspected. The aviation department occasionally hires external inspectors to help its internal personnel perform pavement inspections. All data collected from the inspections is processed and reported to the airport manager who compares it to the data from previous inspections and requests any necessary corrective measures. Table 3-10 summarizes the current tasks of each of the three IG departments.
42

Table 3-10 Current IG Tasks

Department of IG
Railway department
Aviation department

Tasks
1. Monthly manual visual observation Walking through the railway--checking conditions taking pictures documenting issues Inspecting railway tracks including wood ties and the conditions of the surrounding environment
2. Use of special truck equipped with camera (30-50 miles per day at an average speed of five mph) 1. Visual inspection (performed manually)
Inspecting runway markings and signs (general condition) and condition of pavement Inspection of tree heights and approach angle around runway Equipment range finder inclinometer and measuring wheel 2. Pavement condition inspection external or internal inspector 3. Data-processing and reporting to airport manager pre-/post-visual data comparison

IG Tasks with UAS Integration Potential The railway department could integrate a UAS with low altitude and long-distance flight capabilities into its inspections of track elements. If equipped with a thermal camera a UAS could also provide a temperature profile of the railway and facilitate the assessment of cracks expansion and contraction of the rails and other issues. UASs could also be used to monitor railway crossings from various perspectives.

With respect to the aviation department s operations a UAS could provide enhanced images of obstacles and cracks on airport runways. It could also be used to verify the accuracy of the approach path providing more accurate and reliable information on the height of the tree line surrounding the airport. In addition a UAS could collect topographic data of runways and/or of airport construction areas with acceptable precision for management applications. This aerial data collection would decrease the work hours required for this task. Aerial photography can also facilitate pre- and post-survey
43

comparisons optimizing progress monitoring of construction work at airports. During the FG session an aviation manager stated that UAS adoption could help solve cost issues associated with inspecting the large number of airports in the department s charge. Table 3-11 presents IG operations that could incorporate UASs.

Table 3-11 IG Tasks with UAS Integration Potential

Departments in IG
Railway department
Aviation department

Tasks
1. Low-altitude and long-distance flight with low speed for UAS inspection 2. Temperature profile development
Thermal camera-based Railway condition railway expansion contraction and cracking Railway crossing area inspection with UAS 1. UAS-based 3D model through photogrammetry Inspect/assess runway pavement conditions (i.e. detect and measure cracks)
and obstructions Airport area topography (reduced work-hours and increased accuracy) 2. Different perspectives (aerial photography) Construction progress monitoring Pre/post-survey comparisons of runway 3. More cost-effective airport inspection with reduced reliance on outdated equipment

3.2.4. Summary of FG sessions and tasks with UAS integration potential
In general UASs can be integrated into the progress measurement site monitoring and inspection tasks of all GDOT s divisions addressed in this study providing 3D-engineered data such as point clouds digital elevation models (DEMs) and orthomosaic maps. Table 3-12 summarizes the results of the data analysis developed in this chapter and lists the tasks with potential for UAS integration in all groups.

44

Table 3-12 Potential UAS-assisted Tasks in all Groups

Group Construction
Bridge Maintenance
Intermodal (Railway)
Intermodal (Aviation)

Potential Operations with UAS Integration
1. Generating 3D models with photogrammetry Erosion control Earthwork measurement (quantification)
2. High-frequency site condition inspection (daily or weekly inspection)
1. Time saving on bridges with tall columns (an upward-looking camera and
illumination is required)
2. 3D modeling with photogrammetry
Detect and measure cracks conduct vertical clearance assessment Develop 3D steel beam model for precision comparison of as-built structures
3. Inspection underneath bridge and on underside of deck using various sensors (e.g.
IR or thermal sensors)
1. Low-altitude and long-distance flight at low speeds for corridor inspection 2. Temperature profile
Thermal camera-based Inspect railway condition expansion contraction and cracking
3. Railway crossing inspection with UAS
1. 3D modeling with photogrammetry Inspect and observe obstructions Inspect/assess runway pavement conditions (i.e. detect and measure cracks) Airport area topography (reduced work-hours and increased accuracy)
2. Different perspectives (aerial photography)
Construction progress monitoring Pre-/post-construction survey comparisons of runways
3. More cost-effective airport inspection with reduced reliance on outdated
equipment

45

4. Field Tests of UAS-Assisted Tasks
As discussed in the previous chapter a series of UAS-assisted tasks were derived from the data collected in the FG sessions. (See Table 3-12.)
4.1. Field Test Design
The proposed field test protocol was developed based on the findings from the FG sessions with the participating GDOT personnel. The proposal included three different types of UAS platforms for use in the field tests 1) an off-the-shelf quad-copter (first platform) 2) a developer-grade UAS (second platform) and 3) a fixed-wing UAS provided by a thirdparty service (third platform). In keeping with the proposed experiment design presented in Figure 4-1 the field tests were developed to employ different platforms for various tasks and work environments.
The construction inspection test performed by the CG would involve all platforms for collecting images and generating 3D models through the photogrammetry process. Since the CG had never used laser scanners in their operations due to costs involved the use of a laser scanner-equipped platform (second platform) was of special interest. This same platform could be used in the BMG s bridge inspection tests. Three main bridge elements were selected as points of interest for the bridge inspection experiments 1) top-deck 2) under-deck and 3) bridge foundation.
All three UAS platforms would be used for the IG s airport inspection tests which involved monitoring the progress of airport facility construction and performing inspections of
46

airport runway conditions and obstructions. Lastly the off-the-shelf quad-copter and the developer-grade UAS were considered for railway alignment and crossing inspections as suggested by the railway team.
Figure 4-1 Proposed Field Test Design
4.2. Field Tests - Aviation Group 4.2.1. Test Site Selection
The aviation group within the GDOT IG provided three possible test locations. Table 4-1 presents details about these options including airport name location stage of construction project (if applicable) expected travel time and distance to locations from the Georgia Tech Campus.
47

Table 4-1 Aviation Group Potential Test Sites

Site (Name)
Habersham County Airport
(Cornelia)
Monroe-Walton County Airport
Roosevelt Memorial Airport

Code IA01 IA02 IA03

Location
Hwy 441 Bypass Cornelia GA 30531 111 Spring Street Monroe GA 30655
5A9 Woodbury GA 30293

Start Date
March 2017
Feb 8 2017
March 2017

Work Description
Earthwork
Airport runway inspection and
obstruction evaluation
Earthwork

Expected Distance to Location (Time)
75 miles (1H45M)
56 miles (1H25M)
70 miles

4.2.2. Site Selection Visits Results
The research team visited the three possible locations to evaluate existing site conditions at the sites. Table 4-2 provides information from each site visit. The team also visited a fourth airport on the way to one of the three suggested locations. Based on the site visits the research team selected two test sites (IA01 and IA03). Figure 4-2 shows the locations and provides characteristics of the airports visited during the selection process.

48

Table 4-2 Site Visit Summary

Site (Name)
Habersham County Airport
(Cornelia)
MonroeWalton County
Airport
Roosevelt Memorial
Airport
NewnanCoweta County Airport

Code IA01 IA02 IA03 IA04

Person in Charge Ray Reed Airport
Manager (FBO 706778-0198) Brenda Reed FBO Austin Hulsey Line Manager Cris Baily City Manager (770-2665406) Cy Nuually Airport Manager (678-7253542) Wallace Berry (334740-1994) Mark Blace (770-7830645) Time McGowin (334703-3984)
John D. Carroll Airport Manager (FBO 770-254-8102)

Operational Controls Handheld radio control
advisory frequency (no air traffic control tower) Operation depends on weather conditions (strong winds)
Handheld radio control advisory frequency (no air traffic control tower)
Flight school and sky diving club
Handheld radio control advisory frequency (No Air Traffic Control Tower)
Handheld radio control advisory frequency)

Scheduling Need to schedule in advance (heavy traffic on weekends)
Need to schedule in advance
Need to schedule in advance
Need to schedule in advance

Figure 4-2 Location of the Potential Test Sites 49

Figure 4-3 shows logistical features at each airport e.g. the location of construction projects offices and taxiways. Figures 4-4 4-5 and 4-6 provide additional images of each airport for reference.

(a) Habersham County Airport

(b) Monroe-Walton County Airport

(c) Roosevelt Memorial Airport

(d) Newnan-Coweta County Airport

Figure 4-3 Logistics features at Each Airport

50

(a) Project Location (Earthwork around taxiway)

(b) Taxiway

(c) Airport Office Figure 4-4 Images from Visit to Habersham County Airport

51

(a) Taxiway

(b) Potential Obstructions

(c) Airport Office

(d) Airport Overview

Figure 4-5 Images from Visit to Monroe-Walton County Airport

(a) Taxiway

(b) Airport Office

Figure 4-6 Images from Visit to Newnan-Coweta County Airport

52

4.2.3. Selected Test Sites and Field Tests Schedule
Two sites were selected for the field tests 1) the Habersham County Airport and 2) the Roosevelt Memorial Airport. Both airports were undertaking earthwork activities in March 2017. Table 4-3 summarizes information on the selected sites. Based on the feedback from involved staff the dates of field tests were as follows
1. Habersham County Airport Thursday May 18 2017 9AM to 1PM. (The back-up plan in case of inclement weather was Friday May 19 2017.) 2. Roosevelt Memorial Airport Tuesday May 16 2017 9AM to 1PM. (The back-up plan in case of inclement weather was Wednesday May 17 2017.)

Table 4-3 Selected Test Sites

Site (Name) Code

Personnel performing tasks

Field Test Work Descriptions

Field Test Schedule

Habersham County Airport
(Cornelia)
Roosevelt Memorial Airport

IA01 IA03

Georgia Tech Building Construction Dr. Irizarry and Sungjin Kim
Georgia Tech Aerospace Engineering Dr. Johnson Kyuman Lee and an UAS Operator (with Control System equipped truck)
GDOT Aviation Division Inspection Personnel (Alan and Joseph)
Aerial Photographer (Rick Dobbins) Airport Facility Managers

(1) Earthwork monitoring
(2) Airport inspections

Tuesday May 16 2017 (9 00 a.m. to
1 00 p.m)
Thursday May 18 2017 (9 00 a.m. to
1 00 p.m.)

4.2.4. Field Test Protocol Aviation Group
The research team developed a protocol for airport inspection tests. (See Figure 4-7.) A total of six distinct UAS platforms were used for three different inspection tasks. The platforms used included the following
53

DJI Mavic Pro (quad-copter) DJI Phantom 4 (quad-copter) Yuneec Typhoon H (hexa-copter) Parrot Disco FPV (fixed-wing) Topcon Sirius (fixed-wing provided by the industry partner) DJI Matrice (developer-customized platform). The inspection tasks tested included the following Runway inspection Construction inspection Obstruction inspection.
Each participant in the field tests was assigned a combination of code numbers reflecting the team to which they belonged the task they participated in and the platform they were testing. This coding system facilitated the subsequent data analysis. Table 4-4 shows the code numbers used in the field tests.
54

Table 4-4 Field Test Code Designation

Resource Task
Platform
Team

Description Runway Inspection Construction Inspection Obstruction Inspection DJI Mavic Pro DJI Phantom 4 Yuneec Typhoon DJI Matrice TOPCON Sirius Parrot Disco GT Research Team (PIC and VO) GDOT Airport Inspector GDOT Airport Project Engineer Industry Partner (Skysight)

Code Number AV01 AV02 AV03 P01_1 P01_2 P01_3 P02_1 P03_1 P03_2 T01 T02_1 T02_2 T03

Note AV (Aviation)
P01 Off-the-shelf P02 Customized P03 Fixed-wing
T01 GT T02 GDOT T03 Industry Partner

As the participants arrived at the test locations they took part in a pre-flight meeting and then set up the ground control station (GCS) for each set of tests. Flights were performed for each inspection task tested (AV01 AV02 and AV03). The GT research team and the representative from Skysight (the industry partner) operated the platforms while the GDOT team inspected the points of interest at the GCS (e.g. runway pavement earthwork and surrounding vegetation among others). All UAS pilots from the GT research team and from the industry partner were Part 107-certified. Four platforms were used in the runway inspection test (AV01) to collect still pictures (including pictures for 3D model development) and infrared imagery. In the construction inspection test (AV02) two platforms were used to collect still images (including pictures for 3D model development). Lastly both quad-copter and fixed-wing platforms were tested for the obstruction inspection (AV03) which checks for visual or physical obstructions in the approach path at the ends of a runway. The fixed-wing platforms were particularly useful for simulating a pilot s point of view when approaching the runway during landing. The industry partner was also involved in the runway and obstruction inspection tests (AV01 and AV03). At the
55

end of each test flight the teams would decide collaboratively whether additional flights and/or changes to the takeoff location would be required. Figure 4-7 presents the field test protocol and Figure 4-8 illustrates the platforms used in the field tests. Table 4-5 provides the main technical specifications and other information on the equipment used. A total of three types of data were collected 1) still pictures (including pictures for 3D model development) 2) infrared imagery and 3) videos. Quad-copter and fixed-wing platforms were used to collect still pictures some of which were processed into orthomosaic maps digital elevation models and 3D point cloud-based models. In fully autonomous flights (enabled by a flight mission planning software application) a fixedwing platform recorded videos of the runway approach paths. Infrared imagery was collected and processed to facilitate inspection of a number of elements of airport infrastructure. Data were collected based on the data collection plans developed for each test site. (See Figure 4-9.)
56

Figure 4-7 Field Test Protocol Aviation Group 57

(a) DJI Mavic Pro and DJI Phantom 4 (Quadcopters)

(b) Yuneec Typhoon H (Hexa-copter)

(c) DJI Matrice (Customized Platform)

(d) Topcon Sirius (Fixed-wing)

(e) Parrot Disco FPV (Fixed-wing) Figure 4-8 UAS platforms used in field tests
58

Table 4-5 UAS Platforms Specifications

UAS Platform

Weight (g)

Flight Time (min)

DJI Mavic Pro

734

27

https //www.dji.com/mavic/info

DJI Phantom 4

1 380

28

https //www.dji.com/phantom-

4/info

Yuneec Typhoon H

1 695

25

http //us.yuneec.com/typhoon-hoverview

DJI Matrice 100

2 431

247

23

https //www.dji.com/matrice100/ (camera)

info

TOPCON

Sirius

2 700

50

https //www.topconpositioning.c

om/mass-data-collection/aerial-

mapping/sirius-pro

Parrot Disco

FPV

750

45

https //www.parrot.com/us/drone

s/parrot-disco-fpv

SenseFly Albris

https //www.sensefly.com/drone/ 1 800

22

albris

Photo (MP) 12.35 12.4 12.4
12
16
14
38

Video

Mission Planning Software

C4K

Pix4Dmapper DJIFlightPlanner

Cost $999

C4K

Pix4Dmapper DJIFlightPlanner

$799

4K UHD

None used

$1 199

4K UHD

Pix4Dmapper DJIFlightPlanner

$3 299 $899
(camera)

1080p FHD

MAVinci

$20 000

1080p FHD

FreeFlgiht

$399

720p HD

eMotion 3 Pix4Dmapper

$10 500

59

(a) Roosevelt Memorial Airport
(b) Habersham County Airport
(c) Habersham County Airport Figure 4-9 Data Collection Plans
60

4.2.5 Data Collection - Airport Group 4.2.5.1. Roosevelt Memorial Airport
The field test at the Roosevelt Memorial Airport was conducted on May 16 2017. The site is located at 9620 Roosevelt Highway in Warm Springs Georgia approximately 70 miles from the Georgia Tech campus. (See Figure 4-10.)
Figure 4-10 Roosevelt Memorial Airport Location and Views The GT research team GDOT personnel in charge of airport inspections one airport facility manager and four Skysight representatives were present at the test. (See Table 46.) Five platforms were used to collect still images infrared images and videos. Photogrammetric processing was used to develop 3D models from the geo-referenced still pictures. Table 4-7 lists the platforms used and the data collected during the field test.
61

Table 4-6 Roosevelt Memorial Airport Field Test Attendees

Name Irizarry Javier Kim Sungjin Lee Kyuman
Hur Jeong Haviland Stephen
Hood Alan Edmisten Colette
Harper Bill Dobbins Rick Not provided Not provided Pynenburg Alfons Pnoullen Gam

Team GT-BC (PI)
GT-BC GT-AE GT-AE GT-AE GDOT-Aviation GDOT-Aviation Skysight Skysight Skysight Skysight Meriwether County FD Airport Facility Management

Role Pilot in Command (PIC) Visual Observer (VO) Visual Observer (VO) Visual Observer (VO) Extra PIC Visual Observer (VO)
Airport Inspector Airport Inspector Pilot in Command (PIC) Person Manipulate Control (PMC) Visual Observer (VO) Visual Observer (VO)
Attendee Facility Coordinator

Table 4-7 Dataset from Field Test

Used Platform DJI Mavic Pro
Yuneec Typhoon H DJI Matrice
TOPCON Sirius Parrot Disco

Collected Data Still Images Video
Still Images Infrared Image
Still Images Sill Images
Video

Amount of Data 82 photos
1 video (2 mins 44 secs)
29 photos 29 photos
143 photos 1233 photos
2 videos (15 mins 12 secs)

Figure 4-11 shows the field test setup including GCS components and the personnel involved. Figure 4-12 shows sample products of the field test conducted at the Roosevelt Memorial Airport. The data collected will be described and assessed in more detail below.

62

(a) Hexa-copter Flight

(b) GCS Setup

(c) Ground Control Unit at GCS (Typhoon UAS)

(d) Ground Control Unit at GCS (SYrius UAS)

(e) GT Research Team s Setup

(f) Industry Partner s Setup

Figure 4-11 Field Test Setup 63

(a) Orthomosaic of Runway Section

(b) Approach Simulation Video and Mission Plan

(c) Still Image

(d) Infrared Image

Figure 4-12 Field Test Product Samples 64

4.2.5.2. Habersham County Airport
The field test at the Habersham County Airport was conducted on May 18 2017. The airport is located at Hwy 441 Bypass in Cornelia Georgia approximately 70 miles from the Georgia Tech campus. (See Figure 4-13.) Table 4-8 lists the field test attendees. Table 4-9 lists the platforms used and the data collected during the field tests.

Figure 4-13 Habersham County Airport Location

Name Irizarry Javier Kim Sungjin Lee Kyuman
Hur Jeong Haviland Stephen
Hood Alan Robinson Joseph
Harper Bill Dobbins Rick Not provided Not provided

Table 4-8 Field Test Attendees

Team GT-BC (PI)
GT-BC GT-AE GT-AE GT-AE GDOT-Aviation GDOT-Aviation Skysight Skysight Skysight Skysight

Role Pilot in Command (PIC) Visual Observer (VO) Visual Observer (VO) Visual Observer (VO) Extra PIC Visual Observer (VO)
Airport Inspector Airport Inspector Pilot in Command (PIC) Person Manipulate Control (PMC) Visual Observer (VO) Visual Observer (VO)

65

Table 4-9 Dataset from Field Test

Used Platform
DJI Phantom 4
Yuneec Typhoon H DJI Matrice
TOPCON Sirius Parrot Disco

Collected Data
Still Image Video
Still Image Infrared Image
Still Image Sill Image
Video

Amount of Data
101 photos 2 videos (4 mins 54 secs)
29 photos 29 photos 660 photos 1533 photos 1 video (5 mins 8 secs)

Figure 4-14 shows the field test setup GCS operation and components and the involved personnel. Figure 4-15 shows sample products from the field test conducted at the Habersham County Airport. The data collected will be described and assessed in more details below.

(a) Hexa-copter Flight

(b) GCS Setup (Skysight)

(c) Airport Inspector wearing FPV device

(d) Ground Control Unit at GCS

Figure 4-14 Field Test Setup

66

(a) 3D Model of Construction Site

(c) Approach Simulation Video and Mission Plan

(b) Orthomosaic of Facility (by Skysight)

(d) Infrared Image

(e) Still Image

Figure 4-15 Field Test Initial Products

67

The last field data collection activity at Habersham Airport took place on July 11 2018. For this last data collection session at an airport facility GDOT personnel had the opportunity of piloting the UAS under the guidance of a research team member who performed the role of PIC. As shown in Figure 4-16 GDOT personnel piloted a DJI Phantom 4 UAS and a Mavic Pro UAS to collect progress photos. The test also involved pre-programmed autonomous flights using the Maps Made Easy application.

(a) Linear flight pattern

(b) Grid flight pattern

(c) GDOT Personnel at UAS controls

(d) GDOT Personnel monitoring autonomous flight mission

Figure 4-16 Data collection session at Habersham Airport with GDOT personnel operating UAS.

68

4.3. Field Test Rail Group
Two field tests were performed in rail infrastructure environments. The first was at a section of rail located at Lovvorn Farm Road in Carrolton Georgia and took place on August 7 2017. The research team performed the data collection flights and used the Pix4D Capture application to perform autonomous flights. Sample products of the first of these field test are shown in Figure 417.

(a) Location of test site

(b) DSM of test site

(c) Orthomosaic

(d) Point cloud of site

Figure 4-17 Sample products of first rail location test

69

The second field test took place at sections of rail infrastructure in Lafayette Georgia on July 10 2018. GDOT personnel performed data collection flights under the supervision of a research team member PIC. As shown in Figure 4-18 GDOT personnel piloted the DJI Phantom 4 UAS to collect progress photos. The test also involved pre-programmed autonomous flight using the Pix4D Capture application.

(a) GDOT Personnel piloting UAS

(b) Rail Test Site Location 2 overview

(c) Rail Test Site Location 2 detail

(d) Rail Test Site Location 2 point cloud

(e) Rail Test Site Location 2 orthomosaic
Figure 4-18 Sample products of second rail location test 70

4.4. Field Test - Bridge Maintenance Group 4.4.1. Data collection at 17th Street Bridge
Three different UAS platforms were used to collect still images and videos during a bridge inspection field test. The platforms were tested for two different inspection tasks deck inspection and super-/sub-structure inspection. Each participant was assigned a combination of code numbers reflecting the team to which they belonged the task they participated in and the platform they were testing. Figure 4-19 shows an aerial image of the location of the test environment. Table 4-10 shows the code numbers used in the field test. In addition to the GT research team five GDOT personnel from the BMG attended the field test. (See Table 4-11.)
Figure 4-19 17th Street Bridge Test Location 71

Table 4-10 Field Test Designation Codes

Resource Task
Platform Team

Description Deck Inspection Super/Substructure Inspection DJI Mavic Pro DJI Spark DJI Matrice GT Research Team (PIC and VO) GDOT BMG

Code Number BR01 BR03 P01_1 P01_2 P02 T01 T02

Note BMG (Bridge Maintenance Group)
P01 Off-the-shelf P02 Customized
T01 GT T02 GDOT

Name Irizarry Javier Kim Sungjin Lee Kyuman
Haviland Stephen
Joshua Cofer Ryan Beasley Charles Blue Dana Mccrary Bob O Daniels

Table 4-11 Field Test Attendees

Team GT-BC (PI)
GT-BC GT-AE
GT-AE
GDOT-BMG GDOT-BMG GDOT-BMG GDOT-BMG GDOT-BMG

Role Pilot in Command (PIC) Visual Observer (VO) Visual Observer (VO) Visual Observer (VO) Person Manipulating Control (PMC)
Bridge Inspector Bridge Inspector Bridge Inspector Bridge Inspector Bridge Inspector

During the deck inspection test (BR01) multi-rotors (P01 and 02) were used to collect still images of the top- and under-deck for subsequent development of 3D models of the deck. These images allowed for the inspection of minor scaling transverse cracks core holes and joint failures. For the super- and sub-structure inspections (BR02) platforms P01 and P02 were used to check for hairline cracks large voids and scrapes or spalls. After finishing the data collection process the researchers used photogrammetric processing to develop 3D models. Table 4-12 summarizes the platforms used and the data collected with each UAS platform. Figure 4-20 shows the field test environment involved personnel and sample products (e.g. high-resolution still images). The data collected will be described and assessed in more detail below.

72

Table 4-12 Dataset from Field Test

Used Platform
DJI Mavic Pro
DJI Spark DJI Matrice

Collected Data Still Image Video Still Image Still Image

Amount of Data 99 photos 4 videos 22 photos 41 photos

Figure 4-20 Image-based Bridge Inspection
A second bridge inspection field test was conducted on July 18 2018 at a location under a bridge over the Chattahoochee River on GA 400. In this field test GDOT Bridge
73

Maintenance personnel manipulated two platforms the DJI Mavic Pro and the Parrot ANAFI which was one of the newer platforms available. During this test elements similar to those observed during the first field test were also inspected including the under-deck area and beam supports. The Parrot ANAFI platform was selected for testing because it has a camera that can face upward a feature that no other UAS platform in the consumergrade market possesses and one that is particularly useful for under-bridge inspection. Figure 4-21 shows sample products from the field test.

(a) Second bridge test site location

(b) GDOT Personnel manipulating UAS

(c) Under beam support spalling observed with ANAFI UAS facing up.

(d) Under deck observation with Mavic Pro UAS

Figure 4-21 Sample products of second bridge inspection test site

74

4.5. Field Test Construction Group 4.5.1. Data Collection at SR11 US-129 Project
In the field tests conducted on August 16 and September 8 2017 four different UAS platforms were used to collect still images infrared images and videos to monitor road construction. Again each participant was assigned a combination of code numbers reflecting the team to which they belonged the task they had participated in and the platform they were testing. Figure 4-22 shows the location of the test site. Table 4-13 shows the code number for each task platform and team involved in the field test.

Figure 4-22 US129 Project Location

Table 4-13 Field Test Designation Codes

Resource

Description

Code Number

Note

Task Construction Progress Monitoring

CG01

CG (Construction Group)

Platform Team

DJI Mavic Pro DJI Phantom 3 Professional Yuneec Typhoon Parrot Disco GT Research Team (PIC and VO) GDOT CG

P01_1 P01_2 P01_3
P02 T01 T02

P01 Off-the-shelf P02 Fixed-wing
T01 GT T02 GDOT

75

The GDOT PE was in charge of performing the construction progress monitoring while the GT team operated the platforms. (See Table 4-14 for a list of field test participants.)

Name Irizarry Javier Kim Sungjin
Beaudry Jeana

Table 4-14 Field Test Attendees

Group Involved GT-BC (PI) GT-BC
GDOT-CG

Responsibility during test Pilot in Command (PIC) Visual Observer (VO)
Project Engineer/Manager

Multi-rotors (P01_1 and P01_2) were used to collect still images for the subsequent development of 3D models of the road construction site whereas the fixed-wing Parrot Disco (P02) was used to collect videos and still images. After finishing the data collection process the research team used photogrammetric processing to develop 3D models of the site. Table 4-16 summarizes the platforms used and the data collected with each platform. Figure 4-23 shows the road construction environment and Figure 4-24 provides examples of products from the field tests. The data collected will be described and assessed in more detail below.

Table 4-15 Dataset from Field Test

Used Platform DJI Mavic Pro
DJI Phantom 3
Yuneec Typhoon (08/16/17) Parrot Disco (09/08/17)

Collected Data Still Image
Still Image
Still Image Infrared Image
Still Image Video

Amount of Data 32 photos (09/08/17) 160 photos (08/16/17) 387 photos (08/16/17 - Road) 680 photos (09/08/17 - Road) 49 photos (09/08/17 - Bridge)
21 photos 21 photos 11 photos 3 videos

76

Figure 4-23 Road Construction Environment

(a) Still Images of Road Construction Project (Phantom 3)

(b) Infrared Image (Yuneec Typhoon)

(c) Still Image (Yuneec Typhoon)

Figure 4-24 Field Test Sample Products

77

A total of 547 geo-referenced images collected from this test were processed into 3D models. Figure 4-25 and 4-26 shows samples of 3D models and orthomosaic maps of the road construction project.
(a) Road Construction 3D Point Cloud
(b) Road Construction Digital Elevation (c) Road Construction Orthomosaic Map Model Figure 4-25 Field Test Sample Products 78

(a) Bridge Construction Point Cloud

(b) Bridge Construction Digital Elevation Model

(c) Bridge Construction Orthomosaic Map

Figure 4-26 Field Test Sample Products

The last field data collection activity at the US-129 project took place on July 23 2018. During this last data collection session at a construction site GDOT personnel had the
79

opportunity to pilot the UAS under the guidance of the research team member acting as PIC. As shown in Figure 4-27 GDOT personnel piloted the DJI Phantom 4 UAS to collect progress photos. This test also involved pre-programmed autonomous flight through the use of the Pix4D Capture application.

(a) GDOT personnel at UAS controls

(b) progress photo capture by GDOT personnel

Figure 4-27 Data collection session at US-129 project with GDOT personnel operating UAS.

80

5. UAS Workshop
In order to disseminate preliminary results from initial field tests and introduce GDOT personnel to UAS technology the research team held a workshop in the Caddell Building on the Georgia Tech Campus on July 18 2017.
5.1 Workshop Attendees
Personnel from the Construction Bridge Maintenance and Intermodal Groups as well as from the HERO public safety and legal departments attended the workshop. The industry partner Skysight Imaging and the project s research implementation manager also attended.
5.2 Workshop Sessions and Topics
The workshop included an overview of the research project an introduction to UAS technology an overview of applicable FAA regulations preliminary findings from the first year of the research project and a UAS hand-on activity. The preliminary research findings include a thorough review of UAS-related research by other State DOTs as well as the results from the FG sessions and initial field tests with the aviation department.
The workshop was structured in three sections 1) presentation of UAS-related information 2) brainstorming session and survey questionnaire completion and 3) a hands-on activity and structured group interview session. In the presentation portion the research team provided a description of the results from the FG sessions and field tests with the aviation
81

department. During this section the researchers also performed a Pix4D use demonstration and presented 3D models obtained from UAS-collected data. Lastly samples of visual data collected from the I-85 bridge re-construction project were also presented and discussed. Figure 5-1 shows the setting of the workshop.
Figure 5-1 Workshop setting After completing the presentation portion of the workshop participants engaged in a brainstorming session on how UAS technology could be integrated into their tasks. The discussions covered several topics related to the technological procedural and legal requirements for UAS integration. A survey form with open-ended questions was provided to workshop participants to collect their feedback on the following topics
equipment needs equipment and software capabilities internal operational changes FAA regulation compliance internal usage policies procedures and permissions
82

flight documentation/report needs training and licensing requirements insurance and privacy issues damage liability.
Figure 5-2 shows the setting of this brainstorming session.
Figure 5-2 Brainstorming Session Setting The final session of the workshop was a hands-on activity conducted outdoors in front of the Caddell Building on the Georgia Tech campus. This activity was followed by a structured group interview. During the hands-on activity participants had the opportunity to fly a UAS platform under close supervision of the GT research team. Two pilots with Part 107 certification supervised the activity demonstrating the tasks involved before participants took part in the activity. Participants were able to launch the aircraft from the grassy area in front of the Caddell Building--where the research team had set up the GCS--and fly it over the Van-Leer Building construction site adjacent to Tech Green. (See
83

Figure 9-4.) Participants were asked to perform the following six tasks 1) taking off and climbing to altitude (100 feet) 2) hovering in place 3) performing flight patterns 4) flying to the construction jobsite (point of interest) 5) taking still pictures and 6) returning to and landing at the home location. To comply with FAA regulations all tasks were conducted below 400 feet above ground level (AGL). A total of 10 attendees participated in this activity. Figure 5-3 shows the setting of the hands-on activity session. Figure 5-4 shows the DJI Phantom 4 UAS that was the platform used for the hands-on activity.
Figure 5-3 Hands-on Activity Session 84

Figure 5-4 UAS platform used for the hands-on activity
Following the hand-on activity a structured group interview was conducted with workshop participants to collect their perceptions about additional technological operational and human factors involved in UAS integration into GDOT tasks. A total of 13 GDOT professionals participated in the group interview (including the 10 participants in the hands-on activity). The entire interview session was recorded and all questions were previously evaluated and approved by the Georgia Tech Institutional Review Board (IRB).
5.3 Results of Group Interview
UAS Platform Type The interviewees were asked to indicate which UAS platform was most suitable for their tasks (e.g. multi-rotor or fixed-wing). Construction managers and airport inspectors stated that both platforms could be used in their tasks (for airport runway as well as road
85

construction inspections). However since road construction environments usually involve large jobsites--often extending over two miles--a fixed-wing platform could be of more benefit to construction progress monitoring tasks than multi-rotor UASs.
Conversely multi-rotors were seen as more suitable for bridge and culvert construction inspections bridge inspections highway emergency operations and traffic monitoring. Such tasks cannot rely on fixed-wing platforms since they require significant room for takeoff and landing. In addition some of these tasks (e.g. bridge inspections) require the platform to collect close-up images which a fixed-wing platform cannot do since it cannot approach structures in the way needed to collect detailed images.
UAS Sensors Participants indicated that infrared cameras and light detection and ranging (LiDAR) sensors could be beneficial to their tasks. They also indicated that a UAS platform equipped with a thermal camera could be used at airports to check airport runway marking conditions and monitor runway lighting operation. Nevertheless the interviewees agreed that the most useful resources could be the 3D models developed from geo-referenced 2D images even given that the accuracy of these 3D models is subject to the capabilities of the platform s camera and GPS sensor.
Data Process and Management System Participants also discussed the need for a data processing and management system to support UAS operations as outlined in Table 5-1. Most groups concurred that all data
86

should be properly stored in a secure server including flight log files. They also agreed on the need for a defined process for performing 3D mapping.

Table 5-1 Data Process and Management System Requirements

Task Environment Construction
Bridge Airport HERO Others

Requirements Requires cloud-based software and employee training on software Defined 3D data-mapping process Automated earthwork measurement from UAS data UAS would monitor many construction processes Capability to continuously map project progress in 3D System ability to provide data access depending on the organizational or
staffing level Defined 3D data-mapping process Be able to handle/share large volume of data Needs to consider different types of bridges sizes and surrounding
environments Needs UAS-based work procedure Liability insurance data retention and flight planning Inspection could take longer because the UAS cannot accomplish all tasks. Requires new operational team requires a certified pilot and visual
observer Cloud-based software able to handle large volume of data Cloud-based software Defined 3D data-mapping process Needs further study about developing software for infrastructure domain Considers insurance liability documentation process and federal law (14
CFR Part 107)

Team Composition Participants agreed on four essential roles for UAS operation teams 1) the pilot-incommand (PIC) 2) the person manipulating control (PMC) 3) the visual observer (VO) and 4) project specialists. (See Table 5-2.) Nonetheless team members could perform multiple roles depending on their capabilities training and experience. The PIC must be an FAA-certified pilot--that is he or she must hold a Part 107 Remote Pilot Certificate) and must have knowledge of flight controls and airspace. The PIC can operate the UAS during missions or directly supervise others flying missions. The project specialist who is
87

primarily in charge of project management or inspection could eventually take the place of the PIC if he or she has been trained and certified. According to FAA regulations the PMC is the team member in charge of handling the sensors platforms and missions during the entire operation (FAA 2016). The visual observer is responsible for making sure the aircraft is at a safe distance from surrounding objects by maintaining a line of sight on it. In addition the VO should be familiar with the work environment and sequence of UAS flights. During the aircraft s flight team members should use two-way radios and hand signals for remote continuous communication. Table 5-2 presents the UAS team roles and their respective duties.

Roles PIC PMC VO PE

Table 5-2 Team Composition Requirements
Requirements Holds the highest level of operational training Needs certification (FAA Part 107 Remote Pilot Certificate) Needs continuous communication with others Assists the PIC in operating UAS hardware and software Crew resource management UAS operation air traffic and flight mission planning Not required under the FAA Part 107 Maintains sight of the aircraft Is familiar with GDOT s field tasks equipment and safety procedures The roles of VO and PMC could be held by one person. With proper training and certification PE may take PIC role. Should be involved in all flight operations with the PIC

Privacy Safety and Legal Issues All operations should consider private property pedestrians and traffic near the flight area. Moreover an emergency response plan must be implemented in case of accident or loss of communication between the operator and the aircraft. Insurance for UAS damage liability should be required. The GDOT UAS operation policy should comply with FAA regulations. Privacy protection measures emergency response plans and insurance requirements
88

should be clearly described in the policy as well. Table 5-3 presents the legal issues in terms of privacy emergency response and insurance.

Table 5-3 Privacy Safety and Legal Requirements

Attribute Privacy Emergency Response
Insurance

Requirements Do not fly over people. Follow current FAA guidelines. Apply existing data management policy to ensure privacy. Emergency response plans Classification of emergency situations and corresponding response measures Provide insurance for GDOT operators as well as third party liability
insurance (GDOT requires contractors and consultants to provide their own insurance). A state equipment coverage system could be used.

Other Relevant Issues Interviewees also emphasized the importance of conducting pre-flight inspections and of an adequate GCS setup. Another point raised related to the involvement of third-party UAS operators who should provide certified pilots equipment and insurance. They also suggested that employers have a legal agreement with GDOT regarding data access and management. Table 5-4 summarizes the group interview results.

89

Airport Bridge Construction

Table 5-4 Summary of Group Interview Results

UAS Platform
Fixed-wing & Multirotor
Multi-rotor
Fixed-wing & Multirotor
Multi-rotor

UAS Sensors
High-accuracy telemetry sensors
Infrared camera LiDAR Infrared camera LiDAR and high-accuracy telemetry sensors

Data Management System
3D data processing system automated earthwork measurement and payment calculation system and cloudbased documentation system 3D data processing and documentation system
Cloud-based documentation system

Team
PIC PE and VO
PIC Bridge Inspector and/or VO
PIC Airport Inspector and/or VO

Infrared camera conventional camera

3D data processing and cloud-based documentation system

PIC and VO

Legal Issues
Certified PIC privacy issues emergency response plan insurance.

Fixed-wing & Multirotor

Infrared camera LiDAR and high-accuracy telemetry sensors

Documentation system compatible with current system

PIC and VO

HERO

Others

90

6. Data Processing
This chapter describes the processing of the visual data collected from the various field test sites. Photogrammetry had been the method deployed to obtain several products for use in the GDOT tasks selected for study. The process is described in detail in this chapter through an example. Lastly samples of products from the various tests are presented.
6.1 Photogrammetry Software Selection
A number of software applications can be used to enable the UAS-assisted tasks tested in this research. In short data collected with the UAS platforms can be processed into graphical representations such as 3D models and orthomosaic maps which in turn allow for the inspection surveying mapping and monitoring of infrastructure among other tasks. Some applications available include the following Pix4Dmapper DroneDeploy Agisoft Photoscan Autodesk Recap and PhotoModeler UAS. Table 6-1 provides information on each application referenced.
91

Table 6-1 Sample of Photogrammetry Software Available

Application

Measurement features

Polylines Pix4Dmapper distance surface
volume

DroneDeploy

Volume crop health roof

Agisoft Photoscan
Autodesk Recap

Coordinate distance area
volume Ortho distances pipe diameters angles snap to
objects

PhotoModeler Volume terrain

UAS

contour

Processing mode
Cloud & Local
Local
Local Cloud Local

Cost (per licensed user)
$1 900 Educational Version
$3 500/year Professional Version
$399 per user/month
$179 Standard Edition one
computer
$300 annually
$3 995 permanent license
$2 075 annually $199 monthly

Product website
https //pix4d.com/wpcontent/uploads/2017/10/
Pix4Dmapper-V4.0Feature-
List_NEW_versionm.pdf
https //prismicio.s3.amazonaws.com/dr
onedeploywww%2Fd25e2331-
b928-471d-9c6bfbbc4f7e456b_dronedepl
oy-pricingcomparison.pdf http //www.agisoft.com/p df/photoscan_presentatio
n.pdf
https //www.autodesk.co m/products/recap/overvie
w
http //www.photomodeler .com/products/UAS/defa
ult.html

Pix4Dmapper was selected as the primary data processing software tool for this research project due to a number of advantages it has over its peers. This application is highly compatible with the DJI equipment used on the project. It provides the option to process data locally or in the cloud (as shown in Table 6-1). It can also automatically separate a dense point cloud into five groups ground road surfaces buildings high vegetation and human-made objects. The application interacts with a companion mobile application that facilitates the planning of autonomous data collection missions. Lastly Pix4Dmapper also has a "floating license" feature that allows for the activation/deactivation of a license tied to a certain computer at any time in case a new computer is required. This provides flexibility to users since the license can be shared by several users.
92

6.2 Photogrammetry Process with Pix4D
The first step in the application of photogrammetry to images collected with a UAS is to design an efficient image acquisition plan taking into account the following factors project purpose and type (aerial terrestrial mixed) type of camera the rate distance and angle at which images are taken and the flight path(s). On aerial projects for instance paths can be of different types corridor regular grid or circular grid. Deciding on whether more than one flight is needed to cover the full area is also critical. In such cases it is important to determine the area to cover on each flight. The second step before starting a project is to configure the camera settings. Wrong configuration can result in images with unwanted blur noise and distortions. (For specific details on camera configuration refer to the software developer s documentation.)
93

Figure 6-1 Data Processing Workflow
Creating a new project on Pix4D involves the following five steps 1) starting and saving a new project 2) importing the images 3) setting up the image properties 4) selecting the output/ground control point (GCP) coordinate system 5) and selecting the processing options template. Because images carry internal geo-location information it is important to define the coordinate system on Pix4D in order to import the geo-location information along with the images. This is especially important when using GCPs.
Lastly when processing a new project the following steps are recommended initial processing analyzing the quality report point cloud and mesh development and Digital
94

Surface Model (DSM) orthomosaic and index processing. Figure 6.1 shows the photogrammetry workflow with Pix4Dmapper and the following sections briefly describe the process steps. Creating a New Project To create a new project start Pix4Dmapper and then click Project New Project on the menu bar. Then type a name for the project and keep the default option New Project selected in Project Type as shown in Figure 6-2.
Figure 6-2 New Project Window 95

Next to import the images click Add Images in the Select Images window. On the Select Images pop-up navigate to select the folder in which the images are stored select the images to be imported (it is possible to select multiple images) and then click Open. As shown in Figure 6-3 the New Project wizard displays the Image Properties window which contains three separate sections for image geolocation the selected camera model and a table of images.
Figure 6-3 Image Properties Window 96

Image Geolocation This function sets the coordinate system to which the image geo-location data refers. It imports or exports coordinates and if needed it registers the orientation of the images and/or the accuracy of the coordinates. It also sets the accuracy of the image geo-location. If the image geo-location information is stored in the EXIF of the images it will be loaded automatically.
Selected Camera Model This function sets and configures the camera model associated with the images. If the software cannot recognize the camera model different camera parameters can be submitted by editing the camera model.
Images Table This section of the Image Properties window displays the selected images and provides the group position position accuracy and orientation of each image. It also registers whether the image is enabled or not. (Only enabled images can be processed.)
In the Processing Options Template window select the desired template (it can be edited or changed before processing). The 3D Maps option is selected by default but many others templates are available. (See Figure 6-4 and Table 6-2.)
97

Figure 6-4 Processing Options Template Window

Table 6-2 Characteristics of Each Processing Options Template

Processing Options Template 3D Maps
3D Models

Characteristics
Generates a 3D map (e.g. point cloud or 3D textured mesh) as well as a DSM and an orthomosaic.
Image acquisition nadir or oblique flight. Typical input aerial images acquired using a grid flight plan with high overlap. Outputs quality/reliability high. Processing speed slow. Application examples quarries and cadasters among others. Generates a 3D model (e.g. point cloud or 3D textured mesh). Image acquisition oblique flight or terrestrial. Typical input any images with high overlap. Outputs quality/reliability high. Processing speed slow. Application examples 3D models of buildings objects ground imagery indoor
imagery and inspection among others.

98

Ag Multispectral
Ag Modified Camera
Ag RGB
3D Maps Rapid/Low Res
3D Models Rapid/Low Res Ag Modified Camera Rapid/Low Res
Ag RGB Rapid/Low Res
Thermal Camera
ThermoMAP Camera

Generates reflectance index (such as NDVI) classification and application maps. Image acquisition nadir flight with multispectral camera. Typical input images from multispectral cameras (Sequoia Micasense RedEdge
Multispec 4C etc.). Outputs quality/reliability high. Processing speed slow. Application examples precision agriculture. Generates reflectance index (such as NDVI) classification and application maps. Image acquisition nadir flight with modified RGB camera. Typical input images taken with modified RGB camera. Outputs quality/reliability high. Processing speed slow. Application examples precision agriculture. Generates an orthomosaic for precision agriculture. Image acquisition nadir flight over flat terrain with RGB camera. Typical input images taken with RGB cameras for agriculture (Sequoia RGB). Outputs quality/reliability high. Processing speed average. Application examples digital scouting report claiming for precision agriculture. Faster processing of the 3D Maps template for assessing the quality of the acquired
dataset. Outputs quality/reliability low. Processing speed fast. Faster processing of the 3D Models template for assessing the quality of the
acquired dataset. Output quality/reliability low. Processing speed fast. Faster processing of the Ag Modified Camera template for assessing the quality of
the acquired dataset. Output quality/reliability low. Processing speed fast. Faster processing of the Ag RGB template for assessing the quality of the acquired
dataset. Output quality/reliability low. Processing speed fast. Generates a thermal reflectance map. Image acquisition nadir flight with thermal camera. Typical input images taken with thermal cameras (such as Tau 2 based cameras
FLIR Vue Pro FLIR XT). Output quality/reliability high. Processing speed slow. Generates a thermal reflectance map. Image acquisition nadir flight with thermoMAP camera. Typical input images taken with a thermoMAP camera. Output quality/reliability high. Processing speed slow.

Finally click Finish to close the wizard and start the project. Once the project is created
the Map View is displayed. (See Figure 6-5.) 99

Figure 6-5 Map View Window Processing To start processing the project first click View Processing on the menu bar. When the Processing bar opens at the bottom of the main window make sure that 1. Initial Processing is selected and that 2. Point Cloud and Mesh and 3. DSM Orthomosaic and Index are not selected. Click Start. (See Figure 6-6.)
Figure 6-6 Processing Bar (Initial Processing) 100

Once the initial processing is completed the quality report is automatically generated. To deactivate its automatic display unselect the Display Automatically after Processing box at the bottom of the Quality Report window. When more than one step is processed in sequence and processing is complete the quality report PDF file is created in the results folder. The following information should be verified in the quality report
Quality Check Make sure all check boxes are green as shown in Figure 6-7. All or almost all images should be calibrated in one block. The relative difference between initial and optimized internal camera parameters should be below five percent. If using GCPs the GCP error should be below 3GSD.
Figure 6-7 Quality Check Preview On projects that require nadir images and for which the orthomosaic
preview has been generated make sure the orthomosaic does not have holes or distortions. If GCPs or image geo-location has been used ensure that it has the correct orientation. (See Figure 6-8.)
101

Figure 6-8 Preview of Orthomosaic and Corresponding DSM
Initial Image Positions If the images have geo-location verify that the Initial Image Positions figure corresponds to that of the flight plan.
Computed Image/GCPs/Manual Tie Points Positions If using images with geolocation make sure the computed image geo-location is good. If using only images with geo-location check that the uncertainty ellipses are similar in size. If using GCPs ensure that their error is low (i.e. that the difference between input and computed GCPs is small). If using GCPs and images with geo-location the uncertainty ellipses should be very small for images close to the GCPs and may increase for images further away.
Absolute Camera Position and Orientation Uncertainties For projects with image geo-location only make sure that the absolute camera position uncertainty
102

is similar to the GPS accuracy and verify that the sigma is smaller than the mean. For projects with GCPs the absolute camera position uncertainties should be similar to the accuracy of the GCPs. (See Figure 6-9.)
Figure 6-9 Absolute Camera Position and Orientation Uncertainties 3D Points from 2D Keypoint Matches Make sure that enough matches have been
computed between the images and that the graph consists of one block. (See Figure 6-10.) If multiple blocks exist each block should have a different color. The uncertainty ellipses should be of approximately the same size throughout the project.
Figure 6-10 Computed Image Positions with Links between Matched Images
103

Geolocation Details If using GCPs make sure they are all taken into account (i.e. none are displayed in red on the Geolocation and Ground Control Points table). Also verify all marked GCPs. (See Figure 6-11.)
Figure 6-11 Verification of Ground Control Points Next to start the Point Cloud and Mesh step click View Processing on the menu bar. When the Processing bar opens at the bottom of the main window select 2. Point Cloud and Mesh making sure that 1. Initial Processing and 3. DSM Orthomosaic and Index are unselected. (See Figure 6-12.) Then click Start.
Figure 6-12 Processing Bar (Point Cloud and Mesh) 104

Lastly to start the DSM Orthomosaic and Index step click View Processing on the menu bar. When the Processing bar (Figure 6-13) appears at the bottom of the main window make sure that 3. DSM Orthomosaic and Index is selected and that 1. Initial Processing and 2. Point Cloud and Mesh are unselected. Then click Start.
Figure 6-13 Processing Bar (DSM Orthomosaic and Index) Once the project has been completely processed it is possible to use the results in many ways and for different purposes. (See Table 6-3.)
105

Optional Using the ray Cloud Using the Volumes

Table 6-3 Uses of the Results
It can be used to Visualize the different elements of the reconstruction (e.g. camera positions
reprojections (rays) GCPs manual/automatic tie points processing area clipping box densified point cloud terrain/objects/other point groups 3D textured mesh video animation trajectories) and their properties. Verify/improve the accuracy of the reconstruction of the model. Visualize point clouds/triangle meshes created in other projects or with other software. Georeference a project using GCPs and/or scale and orientation constraints. Create orthoplanes to obtain mosaics of any selected plane (e.g. building facades). Assign points of the point cloud to different point groups. Improve the visual aspect. Create objects and measure distances (polylines) and surfaces. Create 3D fly-through animations (video animation trajectories). Export different elements (GCPs manual/automatic tie points objects video animation trajectories). Export point cloud files using points belonging to one or several classes.
Draw volumes. Measure volumes. Exports the measurements.

Using the Mosaic Editor
Using the Index Calculator
Uploading Project Files

Visualize the DSM (raster GeoTIFF digital surface model). Visualize the orthomosaic. Improve the visual aspect of the orthomosaic.
Generate an index map/index grid on which the color of each pixel is computed using a formula that combines different bands of the reflectance map(s).
Provide information about the bands of the reflectance map(s) and index map. Visualize the index map as a colored index map by applying a color mapping to it. Export a georeferenced colored index map. Annotate the classes of the index map to generate an application map. Export an application map as a shape file to be imported in any tractor consoles. Upload the reflectance map on MicaSense Atlas platform. Upload files to the Pix4D Cloud in order to o Store files in the Pix4D online account. o Process projects online. o Provide project information to the support team. o Upload 3D textured mesh to Sketchfab for viewing interacting and sharing.

Using output files in other software

Pix4Dmapper outputs are compatible with many other software applications (GIS CAD etc.) and can be used for many different purposes.

106

7. Data Analysis
Upon completion of the field tests and the subsequent data processing the GT research team discussed the products with the GDOT personnel and industry partners involved in the study. (See Figure 7-1.) These collaborative discussions took place in debriefing sessions on December 11 and 15 2017 and January 24 2018 and consisted of three main steps
1) Short description and discussion of the field-testing outcomes 2) 2D and 3D data demonstration 3) Structured follow-up interviews and survey.
7.1 Data Analysis Structure and Instruments
A total of 12 GDOT professionals participated in the structured follow-up interviews and survey during the collaborative data analysis sessions. Table 7-1 presents the demographic information collected from these sessions. Participants were asked to assess their familiarity with the technologies investigated or with other UAS platforms and 3D computational models. (See Table 7-1.) The ultimate goals of the interviews and survey questions were as follows
1) To identify and classify the performance factors affecting UAS integration (Section 7.2)
2) To develop a conceptual UAS-based workflow (Section 7.3) 3) To collect participant perceptions on the usefulness suitability or adequacy of the
components of the UAS-based workflow to the tasks they perform (Section 7.4)
107

4) To provide future UAS operators at GDOT with insight into how and to what extent a UAS can help them achieve their task-related goals.

(a) Debriefing meeting with Intermodal Group

(b) Debriefing meeting with Construction Group

(c) Debriefing meeting with Bridge Maintenance Group Figure 7-1 Debriefing Sessions Setting
108

Table 7-1 Demographic Information of Debriefing Session Participants

Attribute Gender Male Female Age Under 30 years 31-40 years 41-50 years Over 51 years Group Construction Group Intermodal Group Bridge Maintenance Group Work experience Less than 10 years 11-20 years Over 21 years Educational Attainment High-school level Undergraduate level Graduate level Familiarity with UAS High level Average level Low level Familiarity with 3D High level Average level Low level No Familiarity

Participants (N 12)
83.3% 16.7%
8.3% 41.7% 25.0% 25.0%
41.7% 33.3% 25.0%
66.7% 16.7% 16.7%
50.0% 41.7% 8.3%
16.7% 25.0% 58.3%
0.00% 33.3% 50.0% 16.7%

7.2 Performance Factors
GDOT professionals who participated in the debriefing sessions were asked to indicate the extent to which the listed factors would affect UAS use and performance in their tasks. Participants used a Likert scale to determine the relevance of each of these performance factors. The values of the scale ranged from 1 (representing "not relevant") to 5 (representing "very relevant"). The ranking data were computed and described as mean values. From the post-field-test interviews the researchers identified the following main performance factors
109

1) Hardware Capability of UAS platforms and computer workstations 2) Usability Ease of use (UAS and software) 3) Time Total operating time 4) Cost Total operational cost 5) Human/Team Capability of UAS operators communication and interaction
team composition inclusion of third-party personnel 6) Data Quality 2D and 3D data quality 7) Legal Safety management emergency response and privacy issues.
The safety of operators and bystanders was identified as the most relevant factor (avg. 5.000). The respondents also identified 2D (avg. 4.909) and 3D data accuracy (avg. 4.273) as critical factors. Data quality was cited as having a significant impact on the performance of UAS-assisted progress monitoring and inspection tasks.
With respect to human and team factors the participants indicated that team composition was the most relevant factor (avg. 4.727). Another important aspect was the capability of operators (avg. 4.364) which usually involves their cognitive and task performance. The cost factor (avg. 4.091) and the capability of the UAS platform (avg. 4.545) were other relevant factors that were seen as considerably affecting UAS performance in GDOT tasks. In addition ease of use (avg. 3.545) and operational time (avg. 3.000) were found to be other significant performance factors and should also be considered for the proper UAS operation. Figure 7-2 shows the participants ranking of the relevance of the performance factors.
110

Figure 7-2 Relevance of Performance Factors
7.3 Conceptual UAS-based Workflow
This section presents a UAS-based workflow that can be integrated into construction inspection and progress monitoring tasks as well as airport and bridge inspection tasks. The total operating time of the developed UAS-based workflow was based on survey respondents estimates of the operating time of each step in each workflow given their experience during the field tests. The estimated total operating time for the UAS-based workflow was then compared to the operating time of the existing workflows to determine their relative efficiency.
The developed UAS-based workflows consist of three main steps 1) pre-flight 2) flight and 3) post-flight. (See Figure 7-3.) The pre-flight stage of the workflow comprises the onsite meeting for pre-data collection and flight mission planning equipment setup and checking. The post-flight step consists of equipment disassembly data processing data analysis (debriefing meetings) and data documentation.
111

Figure 7-3 UAS-based Workflow vs. Existing Workflow
Pre-Flight Stage The pre-flight stage consists of four main steps 1) onsite meeting 2) GCS set up 3) GCP set up (if needed) and 4) equipment inspection. The main objective of the pre-flight stage is flight preparation which involves establishing flight objectives deciding on the points of interest (specific locations) developing a mission plan based on those points and setting up the equipment (e.g. GCS GCP or other components).
The main goal of the onsite meeting is to determine flight specificities such as the takeoff and landing locations (also alternate landing locations) potential obstructions and points of interest. The outcome of the meeting should be a detailed flight mission plan. The participants of this meeting should include the PEs and UAS operators.
The GCS should be properly and safely installed somewhere in the work environment (e.g. jobsite airport or bridge). The GCS includes the UAS control system the operators communication system backup batteries and other equipment to support UAS operations as needed. These items may vary depending on the site location and type of project or work
112

environment. After the GCS is set up the UAS and supplementary equipment must be reinspected to make sure the platform is ready to fly.
UAS operators including the pilot-in-command (PIC) and the visual observer (VO) must maintain direct communication during the pre-takeoff checks before starting the flight mission. This can be accomplished with the use of a two-way radio. This is one of the most important steps to take to avoid non-compliance with many important mission parameters required for a safe flight. Performing the pre-takeoff check can prevent accidents and connection loss during flight. For example if the pilot neglects to check available battery power the mission could be affected by sudden power failure. The checklist items include very specific and simple tasks such as checking the UAS battery charge and camera status among others. Figure 7-4 illustrates the pre-flight stage during one of the field tests.

(a) Pre-flight Meeting

(b) GCS Setup Figure 7-4 Pre-flight Stage

(c) GCP Installation

113

Flight Stage (Data Collection) During flight VOs are responsible for visually tracking aircraft position so that pilots can focus on flight control and collection of visual assets. VOs must also check flight conditions using a mission checklist. Three main points are critical during the flight stage 1) confirmation that the aircraft is under the pilot s control and that the GPS has engaged (by hovering approximately 10 feet above ground immediately after takeoff) 2) verification that all control sticks operate correctly while in hover mode and 3) certainty that battery charge levels are safe for flight. The aircraft router (if used) and the transmitter should be at adequate levels of charge (above 50 percent) to prevent connection loss during flight. Flight duration should be kept to no more than 15-20 minutes depending on the UAS platform in use and its battery capacity. When the pilot determines that a mission has been completed a signal should be made to the VO to prepare for landing. The VO then checks on landing location conditions. The landing location should be the same as the takeoff location as determined during the pre-flight planning meeting. If the location is clear and safe for landing the VO sends an approval signal to the PIC. Then the PIC must verify that the camera provides a clear view of the landing location and then proceed to land. If landing location is not available due to obstructions or other issues an alternate landing location can be used. Figure 7-5 illustrates the flight stage during one of the field tests.
114

(a) Takeoff

(b) Data collection

Figure 7-5 Flight Stage

Post-Flight Stage The post-flight stage comprises the data processing tasks. The time needed for data processing depends on the number of images collected the resolution of the images and the specifications of the software application used. Data can be directly downloaded to local storage media or it can be transferred to web-hosted storage which can take from a few minutes to several hours. The variability of web-hosted storage is a function of the bandwidth of communication networks and depends on the read-write speeds of the storage devices used.

In this study local storage was used since it is the most efficient data transfer method. Web-hosted storage is recommended for backup purposes and for non-time-sensitive data sharing. Once data are downloaded processing involves cataloging the visual assets collected by location and work task. This step may require significant time since it is completely manual. Once data has been transferred it can be processed according to the task needs of the various groups.
115

Once the images have been processed stakeholders can discuss evaluate and use the results for their tasks. Should another flight be required proper takeoff and landing locations can be determined as well as points of interest to be inspected.
Estimated Operating Time Survey participants were asked to estimate the operating time of each step of the proposed workflow as well as of the existing workflow for the tasks they perform. The most timeconsuming steps of the UAS-based workflow were identified as the pre-flight setup of the GCS and supplementary equipment and post-flight data processing. The respondents estimated that the GCS and equipment setup take an average of five hours whereas data processing was estimated at three hours. Indeed since data processing is automated it can be performed after business hours with project personnel able to leave the computers and software to run the photogrammetry process by itself overnight if necessary.
In contrast the most time-consuming step for both the existing task methods and a UASintegrated method would be the placement of GCPs when needed. This particular step would apply to tasks that require precise location data such as for construction monitoring tasks (e.g. volume calculations or elevation and distance measurements) some airport inspection tasks for which measurements are needed and bridge inspection tasks that require precise elevation data. GDOT personnel estimated that in most situations it could take approximately 10 hours to establish GCPs. Therefore the total operating time of a flight will vary depending on whether a new GCP is needed for each flight mission. This
116

is to say that the UAS-based workflow can include an existing GCP layout or may require a new one that would require more operating time.
In summary the UAS-based workflow offers significant improvements to data collection and analysis. Based on time estimates UAS-based inspection of a given construction site airport or bridge would take an average of 0.42 hours whereas with the existing method the inspection of the same jobsite or location would take 1.83 hours. In the UAS-based workflow stakeholders take 0.5 hours to perform data analysis and make decisions. By contrast with the existing method data analysis by stakeholders would require an average of 3.53 hours. With respect to the total operating time an average of 11.92 hours was estimated for the UAS-based workflows--assuming that a new GCP layout is not required or that GCPs are already installed. With the existing method the total estimated operating time was an average of 18.075 hours. However if a new GCP setup is needed for the UASbased data collection the total time would increase and the UAS-based integrated method would take longer than the existing method. Table 7-2 shows estimated operating times for the tasks considered in both workflow scenarios.
117

Table 7-2 Estimated Workflow Operating Times

Workflow step
1. Onsite meeting (pre-data collection flight mission planning)
2. GCS setup and installation 3. GCP setup and installation 4. Equipment setup 5. Data collection 6. Data processing 7. GCS removal 8. Data analysis and decision-making 9. Data documentation
Total estimated operating time

UAS-based method (hour)
0.500
1.000 0.000 4.000 0.420 3.000 1.500 0.500 1.000 11.920

Manual method (hour)
0.042
0.000 10.000 0.000 1.833 1.750 0.000 3.533 0.917 18.075

7.4 Usefulness/Suitability Analysis
During the de-briefing meeting GDOT participants were asked to complete a survey designed to assess their perceptions of the usefulness/suitability of identified performance factors to their tasks. Specifically the survey asked how useful/suitable a UAS-based method would be to each task based on performance factors described in Section 7.2. The purpose of the analysis is to better understand the implications of UAS integration into GDOT operations in terms of data collected teams performing the data collection the changes to workflow and the safety of the process. The usefulness/suitability of each factor was based on the participants subjective perceptions of the UAS-based workflow during the field tests and in the debriefing sessions. A Likert Scale with values ranging from 1 (representing "not useful/suitable") to 5 (representing "very useful/suitable") was used. Mean rating values are used to describe results.

118

7.4.1. Usefulness/Suitability of Visual Data (2D and 3D)
Participants in the debriefing sessions were presented with the results of data processed through the photogrammetry software Pix4D as well as with the images and videos collected during the field tests. The following sections present the results for each of the groups that participated. Participants from each group completed the provided survey and rated the usefulness/suitability of the presented UAS-based visual data related to their tasks. This visual data included 2D still images infrared images videos and orthomosaic and point clouds (3D) obtained from the photogrammetric processing.
7.4.1.1 Construction Group Participants indicated that 3D data are highly useful/suitable (avg. 5.000) to construction inspection and progress monitoring. 2D still images are also highly useful/suitable (avg. 5.000) to all tasks (i.e. construction progress monitoring site inspection and surveying). Infrared images were rated as not useful/suitable to any of the CG s tasks (avg. 1.000). Figure 7-6 presents a histogram of the CG s ratings of the usefulness/suitability of the visual data (2D and 3D data) to its tasks.
Figure 7-6 Usefulness/suitability of Visual Data (CG) 119

In terms of best viewpoints to capture useful data CG participants indicated close-up viewpoints were highly useful/suitable for progress monitoring (avg. 5.000) and survey tasks (avg. 5.000). On the other hand site condition inspections would need to be performed from high altitude vantage points (avg. 5.000). Depending on the task scope and goal the flight mission should be adjusted so that the PIC is able to collect high-quality visual data from the best-suited viewpoints. Figure 7-7 shows the best-suited viewpoints to CG s tasks.
Figure 7-7 Best-suited Viewpoints for Data Collection (CG) 7.4.1.2 Intermodal Group The participants in the Intermodal Group (IG) indicated that still images were not very useful/suitable to the assessment of runway approach slope to threshold (avg. 2.000). Infrared images were also rated as not useful/suitable to any airport inspection-related task (avg. 1.000 or 2.000). Videos were rated as relatively useful/suitable (avg. 3.000) to the assessment of runway lighting conditions wind
120

indicator operations and threshold and fueling area condition assessment. Figure 7-8 shows the IG participants ranking of 2D data usefulness/suitability to their tasks.
Figure 7-8 Usefulness/suitability of 2D Data (IG) IG participants indicated that 3D data were highly useful/suitable (avg. 5.000) to the assessment of runway and taxiway design issues runway surface components or safety areas runway approach slope to threshold and runway centerline-related items. IG participants indicated that close-up viewpoints were very useful/suitable for the inspection and monitoring of runway surface components or safety areas wind indicator operations runway lighting threshold and fueling area conditions (avg. 4.000). On the other hand the inspection of runway and taxiway design issues require medium to high altitude vantage points (avg. 4.000). Figure 7-9 shows the best-suited viewpoints for IG tasks.
121

Figure 7-9 Best-suited Viewpoints for Data Collection (IG) 7.4.1.3 Bridge Maintenance Group Participants of the Bridge Maintenance Group (BMG) rated still images as very useful/suitable to the inspection of deck core holes structure debris and large voids (avg. 4.000). Infrared images on the other hand were rated as not useful/suitable to any bridge inspection-related task (avg. 1.000 or 2.000). Video images were also rated as very useful/suitable (avg. 4.000) to the assessment of structure debris exposed footing and large voids. Figure 7-10 shows the usefulness/suitability of 2D data to BMG tasks.
Figure 7-10 Usefulness/suitability of 2D Data (BMG) 122

Participants also indicated that 3D data was highly useful/suitable (avg. 5.000) to the assessment of structure erosion debris exposed footing and large voids. The participants from all three groups were also asked about the best viewpoints to capture useful visual data. Three options were provided detailed close-up view medium altitude view and high-altitude overview. According to BMG participants close-up viewpoints were highly useful/suitable for the detection of deck transverse cracks core holes structure cracks and corrosion (avg. 5.000). Conversely high altitude vantage points were not considered useful/suitable for bridge inspection tasks (avg. 1.000 or 2.000). (See Figure 7-11.)
Figure 7-11 Best-suited Viewpoints for Data Collection (BMG)
123

7.4.2. Adequacy/Suitability of Team Composition
GDOT participants were asked about the adequacy of team composition for UAS operations based on their field-test experience. In general the operational team would consist of a PIC VO(s) PE(s) and an FC (facility coordinator). The respondents indicated that such a team composition was not very suitable to the development of flight missions (avg. 1.550) or to the performance of data collection (avg. 2.180) or data processing (avg. 2.090). However the same team composition was rated as very adequate for the performance of post-flight data analysis (avg. 4.180). In general GDOT personnel indicated that the team composition would allow for effective decision-making and data management/documentation (avg. 4.180). Figure 7-12 shows the adequacy of the team composition by workflow phase.
Figure 7-12 Adequacy of Team Composition
124

7.4.3. Adequacy/Suitability of UAS-based Workflow
Participants from all groups were also asked to evaluate the adequacy of the UAS-based workflow i.e. to indicate how suitable it was to UAS operations. They rated data collection as the most challenging step of the workflow (avg. 3.200) as shown in Figure 7-13. Handling the UAS platform requires training and to some extent is subject to a pilot s cognitive ability. On the other hand the UAS-based workflow was rated as very suitable to visual data processing data analysis and decision-making (avg. 4.000). These findings suggest that as pilots receive more training on the UAS platform to better perform data collection the UAS-based workflow would become better integrated into GDOT operations.
Figure 7-13 Adequacy of UAS-based Workflow
7.4.4. Efficiency of UAS-based Workflow
Participants from all groups rated the efficiency of the developed UAS-based workflow in contrast to the existing workflow (with no UASs involved). Participants agreed that the
125

UAS-based workflow improves data collection (avg. 4.630) analysis (avg. 5.000) and documentation (avg. 4.603). However 2D and 3D data processing were found to be inefficient when compared to the existing method (avg. 2.000 and 1.380). Indeed as discussed above these tasks take longer with the UAS-based workflow than with the existing method. These results are shown in Figure 7-14.
Figure 7-14 Efficiency of UAS-based Workflow
7.4.5. Safety Considerations
Safety is critical during the pre-flight and data collection flight stages of UAS operations. In some cases it is necessary to perform pre-operation flights to determine the appropriate altitude for data collection distance to points of interest and potential obstacles to safe operations. The participants all said that both public safety (avg. 2.200) and UAS operator safety (avg. 2.600) should be carefully considered. Data collected during the pre-flight phase--from dry runs or "pre-ops" flights--will allow the UAS team to develop a flight mission plan that considers the UAS team safety public safety and privacy of the general public. Personnel indicated that safety on data collection flights would be enhanced
126

as result of this approach (avg. 4.600 or 4.500). Figure 7-15 shows this perceived improvement in safety performance and effectiveness.
Figure 7-15 Safety Improvements (Public and UAS Team)
7.4.6. Summary of Analyses
Mean values were calculated for all the components analyzed as follows relevance of safety procedures (avg. 3.850) adequacy of the team composition (avg. 4.090) efficiency of the UAS-based workflow (avg. 3.250) adequacy of the UAS-based workflow (avg. 3.700). Findings suggest that UAS operations require a team composition suitable for analyzing visual data and effective decision-making. However further studies are required to improve team effectiveness during flight mission development and data collection. With improved team composition and communication the team would be able to improve data quality as well as the efficiency of work procedures. Figure 7-16 shows the overall effectiveness of UAS operation in terms of performance factors.
127

Figure 7-16 Summary of Analyses 128

8. Legal Safety and Privacy Considerations
The following considerations support long term safe and effective UAS integration into the GDOT operations studied. These can be taken as recommendations for accomplishing this objective. Any future UAS operational policy should comply with FAA regulations. Privacy measures emergency response plans and insurance standards should all be clearly described in this policy. The 14 CFR Part 107 rules have established a useful overall structure for safe and efficient UAS integration into GDOT operations. Additional considerations are provided below.
Safety Considerations All operations should pay attention to private property pedestrians and traffic surrounding the flight area. An emergency response plan must be put into effect in case of accidents or loss of communication between the operator and the aircraft. The emergency response plan should include a classification of emergencies and corresponding contingency measures. Also the plan should be provided to the operation personnel in hard copy format and made available for consultation at the GCS.
As much as possible GDOT should educate the public about the characteristics and risks of UAS operations including the aircraft and sensors used the goals and types of flight missions the type of data collected as well as the risk mitigation measures and emergency procedures.
129

Adequate UAS operator training is not only legally required but essential in a practical sense. Nonetheless unforeseen circumstances and accidents may happen and in order to make up for eventual harm or prejudice to victims insurance for UAS damage liability would be needed at minimum coverage levels.
Privacy Considerations UAS operations should avoid flights over private property pedestrians (107.39a) and traffic (107.39b)--not only because of safety issues but also due to privacy matters. Any future GDOT UAS operational policy should draw from the Health Insurance Portability and Accountability Act (HIPAA) to develop protective mechanisms for personal information of victims in the event of accidents.
Legal Considerations By law in order to fly within the U.S. National Airspace System (NAS) an aircraft must be registered the pilot must be certified and the operation may need to get prior approval depending on the type of operator (hobbyist vs. commercial). The FAA requires that hobbyist UAS operation within a five-mile radius of an airport must be coordinated with the airport operator. The hobby UAS operators must communicate with the airport or heliport administration and obtain permission to fly if the airport/heliport facilities are located within five miles of the UAS flight area. In GDOT s case operations are considered commercial and therefore may require prior authorization when operating within five miles of an airport depending on airspace classification. Prior authorization would be required if operating in controlled airspace including Class G controlled airspace. UAS operations
130

also require visual observers to keep track of the aircraft during flight (it must stay within his or her VLOS) to avoid unforeseen circumstances and accidents. Given these considerations for UAS operation GDOT would need to choose one of the following options 1) hire a certified pilot on a contract basis or 2) provide personnel with adequate training and certification.
The rules for non-recreational UAS operations enforced by the FAA are included within Title 14 of the Code of Federal Regulation (14 CFR) Part 107. (See section 8.1 below.) In order to fly UASs that weigh 55 pounds. or more operators would need to go through the FAA Section 333 exemption process (See section 8.2 below). For governmental entities a public Certificate of Waiver or Authorization (COA) is required to fly UASs. (See section 8.3 below.)
8.1 Title 14 of the Code of Federal Regulation Part 107
In order to adhere to the FAA rules regarding small UAS operations an operator must be aware of and meet various requirements (14 CFR Part 107). (See Table 8-1 below.)
131

Table 8-1 Main Requirements of 14 CFR Part 107

Work/Business purpose flights

Must be at least 16 years of age

Must pass an initial aeronautical knowledge test at an FAA-approved

Remote Pilot Certification

testing center Must undergo Transportation Safety Administration (TSA) security

screening

Must pass a recurrent aeronautical knowledge test every 24 months

Must weigh less than 55 lbs. including payload at takeoff

Must be registered if over 0.55 lbs.

Aircraft Requirements

Must be registered under Part 107 if unmanned aircraft not flown under section 336

Must undergo pre-flight check to ensure that UAS is in condition for

safe operations

Location Requirements Fly in Class G airspace*

Must keep the aircraft within visual line-of-sight (VLOS)*

Must fly under 400 feet*

Must fly during the day or civil twilight*

Operating Rules

Must fly at or below 100 mph* Must yield the right of way to manned aircraft*

Must NOT fly directly over people*

Must NOT fly from a moving vehicle unless in a sparsely populated

area*

* Part 107 Sections Subject to waiver Operation from a moving vehicle or aircraft ( 107.25) Daylight

operation ( 107.29) Visual line of sight aircraft operation ( 107.31) Visual observer ( 107.33) Operation

of multiple small unmanned aircraft systems ( 107.35) Yielding the right of way ( 107.37(a)) Operation

over people ( 107.39) Operation in certain airspace ( 107.41) Operating limitations for small unmanned

aircraft ( 107.51)

When operating a UAS pilots also need to consider the airspace classification in the area of operations. It is particularly important to determine whether flights are required within controlled airspace. According to the Aeronautical Information Manual a controlled airspace is defined as "an airspace of defined dimensions within which air traffic control service is provided to both Instrument Flight Rules (IFR) and Visual Flight Rules (VFL) flights in accordance with its classifications" (FAA 2016). In the United States the controlled airspaces are designated as in Table 8-2.
132

Table 8-2 Designated Airspaces in United States (Adapted from FAA (2016))

Airspace Class

Definition

Class A

From 5 500m (18 045 ft.) mean sea level (MSL) up to and including Flight

Level (FL) 600.

Class B

From the surface to 3000m (9 842 ft.)MSL.

Class C

From the surface to 1 200 m (4 000 ft.) above the airport elevation.

Class D

From the surface to 760 m (2 493 ft.) from the airport elevation.

Class E

An airspace that is not classified as A B C and D

Class G

Uncontrolled airspace with no IFR operation.

1.Flight Level (FL) is defined as a nominal altitude in hector-feet while being a multiple of 500-ft. FL 600

is equal to 18 200 m (60 000-ft.)

Some 14 CFR Part 107 rules provide option waivers which allow for a small UAS operation to deviate from certain operating rules should the FAA find that the proposed operation could be performed safely. The certificates of waiver may include special provisions designed to ensure that the small UAS operation offers a level of safety equivalent to that stipulated by Part 107 rules.

8.2 Section 333 Exemption Aircraft weighing more than 55 pounds
The 14 CFR Part 107 rules discussed above are only applicable to unmanned aircraft that weigh up to 55 pounds at takeoff. In order to fly a UAS that weighs 55 pounds or more operators would need to go through the FAA Section 333 exemption process. In this case operating rules and aircraft requirements are identical or similar to small UAS rules. The FAA determines the pilot requirements for the 333 exemption petitions on a case-by-case basis.

133

The Section 333 of the FAA Modernization and Reform Act of 2012 (FMRA) grants the Secretary of Transportation the authority to determine whether an airworthiness certificate is required for a UAS to operate safely within the NAS.
8.3 Certificate of Waiver or Authorization
To legally operate a UAS governmental entities and organizations (e.g. state governments law enforcement agencies public universities and local municipalities) must meet one of the following requirements
Fly under the small UAS rule--adhere to the rules in 14 CFR Part 107 including aircraft and pilot requirements. (See Section 8.1.)
Obtain a public Certificate of Waiver or Authorization (COA) that allows for nationwide flights within the Class G airspace at or below 400 feet selfcertification of UAS pilots and the option to obtain emergency COAs (e-COAs) under special circumstances.
A Certificate of Waiver or Authorization (COA) is a permit issued by the Air Traffic Organization to a public operator for a specific UAS operation. The COA application form requires the following information concept of operation and type of mission operation location altitude communications and flight procedures. After submission the FAA conducts a comprehensive operational and technical review of the application to ensure that the UAS can operate safely with other airspace users. As of 2018 the wait time for application review is 90 days. The COA application also requires proof of the airworthiness of the UAS. This proof can be obtained either by submitting an Airworthiness Statement
134

or through the FAA Certificate of Airworthiness. More recently the FAA has begun to implement the Low Altitude Authorization and Notification Capability (LAANC) which facilitates access to controlled airspace near airports through near-real-time processing of airspace authorizations below approved altitudes. Requests for access can be made through mobile applications from approved service providers.
8.4 State UAS Laws
Several state general assemblies or state legislatures have developed their own laws and regulations concerning UAS operation. Appendix ee provides a compilation of state laws concerning UAS use. In Georgia all commercial UAS operations are subject to the 14 CFR Part 107 rules. Appendix ff presents excerpts from a Georgia state law on UAS operation. It is important to note that there is on-going litigation in many states over questions of whether federal law supersede local laws aimed at regulating UAS operations.
135

9. Recommendations for UAS Integration Guidelines
This chapter will discuss recommendations for UAS integration at GDOT based on lessons learned from tasks considered in the research project. The recommendations provided consider FAA regulations that were applicable during the field tests and at the time of this writing.
9.1 Applicable Regulations Affecting GDOT UAS Integration
Regulations applicable to GDOT UAS operations are discussed in detail in Chapter 8. These regulations put the following limitations on certain aspects of UAS use
The FAA mandates that the PIC maintain line-of-sight with the vehicle during flight. However one of the advantages of using UASs is to gain access to locations that are difficult to reach e.g. bridge elements or distant points on runway or in road construction zones. Maintaining line-of-sight becomes difficult for certain terrain and topographical situations severely limiting inspection abilities. It may be possible to obtain a waiver for these situations but the time required for the approval of such waivers may limit the practicality of using a UAS for the task. Since developing technologies are expected to meet the line-of-sight requirements in the near future they will soon cease being an issue for GDOT UAS operations.
Current FAA regulations prohibit UASs from passing over traffic requiring lane closures. Waivers for flight over traffic are possible but the proximity to traffic is the deciding factor. Again as indicated above the timing for waiver approval could 136

be an issue. However it is possible to mitigate this risk by developing flight plans that collect imagery at oblique angles that do not require passing over traffic. FAA rules limit UAS use on tasks that can benefit from the use of thermal images e.g. deck delamination detection through thermal inertia (which requires taking thermal images of a surface in two different ambient temperatures with maximum possible temperature gradient i.e. daytime and nighttime). Indeed according to the FAA flights are limited to daytime operation. Waivers for nighttime flights are possible but pose the same challenges as indicated above. According to FAA regulations the maximum flight altitude is 400 feet Therefore it would be impossible to inspect any structures exceeding this altitude. This requirement does not affect the inspection of most bridges in the GDOT purview. However lower altitudes reduce the area captured in images and the number of images required to cover larger horizontal areas. Again a waiver is possible to work around this restriction but for many GDOT applications it may not be practical.
9.2 Operational Considerations
In addition to regulatory requirements that could affect UAS use for the GDOT tasks considered in this study there are other issues to consider for UAS operations. The following are recommendations related to various operations-related topics.
137

UAS Operations Planning These recommendations apply to a broad range of GDOT tasks that could benefit from UAS integration beyond the ones considered in this study. Before a UAS operation is performed the following steps could be implemented
Any GDOT employee intending to employ a UAS and taking on the role of PIC should develop a flight plan that includes at a minimum the following information o An airspace review that identifies the class of airspace in which the operation will take place. This will help the PIC determine whether a waiver will be required. Special consideration should be given to locations at or close to operating airports. In case of close proximity to such facilities the PIC should arrange for a Notice to Airmen NOTAM to be issued by the relevant party. In the case of GDOT the Intermodal Group could assist. It is also important that the PIC verify whether there are active temporary flight restrictions (TFRs) in effect that may include the intended operational area. There are many mobile applications that PICs could use to aid in the airspace review e.g. the FAA s Before You Fly app and AIRMAP among others. (See Figure 9-1.) AIRMAP has the benefit of being a Low Altitude Authorization and Notification Capability (LAANC) provider and would allow PICs to make Controlled Airspace Authorization requests and receive instant authorization where available.
.
138

FAA Before You Fly (B4U Fly) Mobile Application

AIRMAP Mobile Application

Figure 9-1 Examples of mobile applications for airspace reviews before UAS operations

o It is critical to obtain forecasted weather-related information for the operating area. Considering weather conditions (e.g. rain snow or thunder storms) and temperature is essential for safe and efficient UAS flight performance. PICs should also consider that per FAA regulations the minimum weather visibility distance is three miles from the control station.
o PICs should provide a diagram depicting the area of operations and should identify any limitations to operations. Takeoff area main landing area and alternative landing areas should be indicated in the diagram. This diagram
139

will facilitate a risk assessment to identify hazards or obstructions to operations. In addition the diagram can be used to identify the need to secure land owner permissions for access during operations. o As part of the planning process UAS pilots should determine whether privacy is a concern and employ reasonable precautions to avoid capturing images of the public except those that are incidental to the project. o The PIC should provide a statement of the purpose of the UAS operation and how it is related to the tasks they will perform. This can be used as a record of the need and purpose of the UAS for the task. This description should include the maximum expected altitude of proposed operation. According to FAA regulations the maximum flight altitude is 400 feet above ground level (AGL). It could be higher if the UAS remains within 400 feet of a structure. o The schedule for the operation should be drawn up to ensure that it meets daytime operation requirements. Flight can be accomplished during daylight or in civil twilight (30 minutes before official sunrise to 30 minutes after official sunset local time) with appropriate anti-collision lighting. If nighttime operations are needed the relevant waiver should be secured and the waiver identifier should be included. o The expected duration of the flight should be calculated. A more detailed value for the duration can be obtained from the UAS telemetry data or from the time stamps of the visual data collected if needed.
140

o It is important to describe the communication plan between the PIC and any VOs involved as well as emergency/contingency procedures in case of incidents e.g. a lost link with the UAS lost communication with the UAS UAS power loss and unexpected emergency landings.
o Important note regarding autonomous UAS Flights A UAS is capable of flying autonomously on GDOT project sites as long as they follow developed UAS flight plans. However autonomous flight requires PICs to monitor the UAS ground control station at all times. They cannot engage in any other tasks during UAS flights. They are also responsible for keeping track of the UAS flight missions with visual observers as needed.
UAS Operation Execution Once the flight plan has been approved by GDOT and the PIC is at a site ready to perform the UAS operation the following recommendations could facilitate the performance of safe UAS operations
A pre-flight checklist should be completed by the PIC. Although software applications are available for completing pre-flight checklists care should be taken to ensure that they meet GDOT record-keeping requirements. According to GDOT s UAS policy published on November 7 2017 all UAS pilots are required to complete a Pre and Post flight report for all UAS flights. Appendix gg includes an example of a pre-flight checklist used by the research team.
Takeoff checklist items could also be used by the PIC by including them in the preflight checklist. Software applications are also available for completing takeoff 141

checklist items but again care should be taken to ensure they meet GDOT recordkeeping requirements. Refer to Appendix gg for an example of a takeoff checklist items used by the research team. Once an operation is completed and the UAS has safely landed a post-flight checklist should be completed. The same considerations regarding the pre-flight and takeoff checklists apply here. After the operation is completed the PIC should copy the data collected including images and video to the GDOT controlled servers. GDOT should apply its chosen method to ensure the security of the data and its accessibility to interested personnel. If a third party (e.g. a contractor) is used to perform UAS operations existing GDOT procedures regarding data transfer should be implemented. During operation the following general flight requirements should be considered
o Battery life The flight must be conducted with enough remaining battery to ensure safe landing at the home point or any other landing point determined on the flight plan and the UAS should have enough reserve battery life to ensure its safe landing at an alternative site if landing at the primary landing site is not possible.
o Flight speed The maximum flight speed is 100 mph (87 knots). Choice of technology for operation GDOT personnel should spend time evaluating
the intended operation to determine which platform is best suited for data collection. During the field tests several platform types were used (e.g. different sizes of multi-rotor and fixed wing models among other) and depending on the task performed some were found to be more useful than others. The experiences
142

chronicled in this report can help GDOT personnel choose the appropriate UAS platform for their tasks. Contracting UAS services If it is determined that an outside vendor is better suited for a particular UAS operation GDOT should ensure the following
o Any UAS service provider selected to perform operations for GDOT should meet existing requirements for consultant services and be able to execute associated policies and procedures.
o Any UAS service provider selected to perform UAS operations for GDOT will follow all GDOT UAS policy requirements additional to any general policies applicable to service providers.
UAS Fleet and Data Management The following are recommendations for the management of any UASs owned by GDOT. They are based on FAA regulations best practices and lessons learned during the field tests of this research project.
o FAA regulations stipulate that any UAS weighing 0.55 pounds or more must be registered with the FAA regardless of type of use (i.e. commercial or recreational). Therefore any GDOT-owned UAS that exceeds the 0.55-pound criterion must be registered. Registration costs $5.00 and can be completed on the FAADroneZone website (https //faadronezone.faa.gov). Once registered all GDOT UASs must display the appropriate markings as required. One GDOT employee should be designated as responsible for UAS registration on behalf of the department and that person should be in a position of authority in any GDOT UAS program. 143

o For GDOT-owned UAS aircraft equipment malfunctions should be noted in the appropriate post-flight checklist.
o All GDOT-owned UAS equipment should be properly maintained according to manufacturer recommendations. In addition to having scheduled annual inspections all UAS equipment should also undergo pre- and post-flight inspections. Any maintenance performed should be documented in maintenance logs as required. Each UAS unit should have its own maintenance documentation. Information that should be required in such forms includes the UAS identification number date of maintenance maintenance performed inspection performed and any necessary additional notes or comments.
UAS Pilot in Command Requirements o GDOT should provide potential UAS pilots with access to training resources on safe UAS operation. Training beyond FAA Part 107 regulations is recommended. Courses offered by UAS pilot ground schools are available from service providers in many locations. These courses help potential UAS pilots understand the National Airspace System (NAS) and learn the rules associated with safe flight within it. This training can prepare individuals to take the FAA Part 107 certification exam in order to obtain the Small Unmanned Aerial System Rating Certificate. o All GDOT UAS PICs or PICs from contracted service providers must possess FAA Part 107 certification to operate UASs on behalf of GDOT.
144

10. Conclusions and Future Research
Unmanned Aerial Systems (UAS) or "drones" are becoming a part of everyday life with a constantly growing impact on the way many tasks are performed. This study aimed to propose guidelines for UAS integration into tasks performed by selected groups within the Georgia Department of Transportation. The guidelines developed from this research are based on the results of several activities focus group sessions with personnel from the department s Intermodal Construction and Bridge Maintenance groups a workshop mainly with personnel from these groups and some from the HERO and Legal groups and field tests in the work environments of the three main groups included in the study. From the focus group sessions seven test sites were identified and tasks were performed with UAS at each site. From the data collected GDOT personnel was able to assess usefulness and usability of the process and the products obtained. The following conclusions are presented based on the results of the activities undertaken in the research study and organized according to the objectives of the study. 1) The first objective of the study was to determine the technological feasibility of
utilizing UASs in the operations of GDOT divisions. GDOT personnel found that in general they could use UAS for the tasks considered in the study. However they recognized that training would be needed to obtain the technical skills required for safe use of the UAS devices. In addition they recognize that in not all situations the best approach would be for GDOT personnel to employ UAS but to have a third party perform data collection with the UAS devices. This was noted in tasks related to airport inspections. It was noted that a third party could perform the tasks in a shorter time and
145

more efficiently given the large area that needs to be covered on each inspection. Another suggestion from GDOT personnel was to have a dedicated group within the department who would provide UAS data collection as a service to the various areas of the department. Thus centralizing all aspects related to the acquisition use and maintenance of UAS at GDOT. In conclusion it was determined that the application of UAS for the tasks considered in the study is technologically feasible provided that GDOT personnel receive proper training. 2) The second objective of the study was to understand the advantages and limitations of UAS adoption (as well as its legal safety and privacy implications) for tasks identified from the analysis of GDOT divisions. The advantages of UAS integration into GDOT tasks considered are clear. Removing personnel from dangerous environments and situations encountered for example when inspecting hard-to-reach locations on bridges and roadways as well as when inspecting airport runways is a significant advantage. The time saved in collecting visual data such as images of issues encountered during the aforementioned inspection tasks is another benefit of the application of UAS. The same visual data can be converted in measurable data such as point clouds when the appropriate software tools are used. GDOT personnel considered the data collected and the results of obtained from processing visual data as useful. The results included 3-dimensional point clouds and orthomosaic images of the facilities inspected and construction projects included in the study. One area of concern was the tools skills and time required to process collected data. GDOT personnel would benefit from training to use the tools needed to process data such as photogrammetry software. Another area of concern related to data collected is storage of data. During
146

the study GDOT personnel discussed options for storage that would leverage existing systems and data maintenance policy the department has in place. In terms of legal and privacy issues it was found that GDOT has mechanisms to protect data collected using current methods available to the department. In terms of liability of UAS operators employed by GDOT it was recommended that GDOT manages such liability to protect its personnel as well as require contractors to provide their own liability insurance as it is standard practice in the department. 3) The third objective of the study was to propose FAA-compatible guidelines for integrating such systems into GDOT operations. The recommended guidelines for the implementation of UAS within various GDOT groups rely on current FAA regulations governing the use of UAS in the NAS. Existing regulations are considered workable in terms of GDOT requirements for the tasks considered in the study. Since GDOT has control to access of most of its work environments management of safety conditions related to UAS operations is possible. This would be particularly beneficial to tasks related to airport inspections and railway inspections where GDOT has strict access control of facilities or right-of-way. However safety precautions are needed when operating in active airports or projects that are in close proximity to the general public as the case of construction work on public roadways. On active airports UAS operations would be preceded by the filing of a Notice to Airman (NOTAM) for the operation period. For UAS operations on roadway projects particularly on expansion re-alignment and other projects involving roadways in use UAS operations would follow recommended guidelines that comply with FAA regulations. A significant concern of GDOT personnel is the liability presented by the use of UAS devices when
147

used in close proximity to the general public. In order to adequately implement UAS for GDOT operations a liability management strategy should be implemented to provide clear guidance to employees regarding liability protection by GDOT while they perform UAS related tasks on the department s behalf. 4) The last objective of the study was to hold a workshop for GDOT personnel about the use of UAS technology for the investigated tasks. The workshop provided personnel with information on UAS technology in general FAA regulations related to UAS software tools for data collection mission planning and software tools for image processing to obtain actionable data and the opportunity experience first-hand the use of an UAS on a simulated task similar to what GDOT would perform at a construction site. Future research on the integration of UAS into GDOT operations should consider the following
Advanced use of data collected with UAS In this study data collected included images videos and infrared images. There are sensors being developed that can collect visual data as well as other environmental data that could benefit GDOT operations. Future research could consider the use of advanced sensors for additional GDOT tasks such as non-destructive inspections of infrastructure monitoring of environmental conditions infrastructure asset management and emergency management tasks. Other potential applications that could be studied would consider the use of machine learning and artificial intelligence to collect and analyze UAS-based data for applications such as traffic management automated
148

verification of contractor work and automated assessment of inspection criteria at facilities such as airport bridges and roads among others. Monitoring of status of implementation As GDOT implements UAS for the tasks identified in the study and others it will be important to track performance of personnel as they implement the technology. Data on location of flights purpose data collected issues encountered use of the data and others will allow GDOT to determine the success of UAS integration into department operations. Use of UAS technology beyond the applications considered in the study In order to explore the full potential of UAS technology for GDOT applications other areas in addition to the ones included in the study should be considered. At the time of this the writing of this report FAA regulations prevent the use of UAS at night or over people or beyond visual line of sight. However these restrictions are bound to be relaxed in the near future providing the department with the opportunity to explore applications such as remote monitoring of construction operations as well as traffic management among others.
149

11. References

AASHTO (2010). AASHTO Bridge Element Inspection Guide Manual

Anand S. (2007). "Domestic use of unmanned aircraft systems an evaluation of policy

constraints and the role of industry consensus standards." ASTM

STANDARDIZATION NEWS 35(9) 30.

Barazzetti L. Remondino F. & Scaioni M. (2010). Automation in 3D reconstruction

results on different kinds of close-range blocks. Paper presented at the ISPRS

Commission V Symposium Int. Archives of Photogrammetry Remote Sensing and

Spatial Information Sciences Newcastle upon Tyne UK.

Barfuss S. L. Jensen A. & Clemens S. (2012). Evaluation and development of unmanned

aircraft (UAV) for UDOT needs. Retrieved from Utah State Department of

Transportation

website

https //www.udot.utah.gov/main/uconowner.gf n 10710706202834543

Blinn N. & Issa R. R. A. (2016). Feasibility Assessment of Unmanned Aircraft Systems

for Construction Management Applications. Paper presented at the Construction

Research Congress 2016.

Brooks C. Dobson R. J. Banach D. M. Dean D. Oommen T. Wolf R. E. Hart B.

(2015). Evaluating the Use of Unmanned Aerial Vehicles for Transportation

Purposes. Retrieved from Michigan Tech Research Institute Ann Arbor Michigan

Bryant S. et al. (2016) UAS Recommendation Report Illinois State Department of

Transportation.

150

Chan B. Guan H. Jo J. & Blumenstein M. (2015). "Towards UAV-based bridge inspection systems a review and an application perspective." Structural Monitoring and Maintenance 2(3) 283-300.
Carroll E. A. & Rathbone D. B. (2002). Using an Unmanned Airborne Data Acquisition System (ADAS) for Traffic Surveillance Monitoring and Management. New Orleans LA United States American Society of Mechanical Engineers 145-157.
Coifman B. McCord M. Mishalani R. G. & Redmill K. (2004). "Surface transportation surveillance from unmanned aerial vehicles. Proceedings of the 83rd Annual Meeting of the Transportation Research Board.
Denning J. D. & Verschelden C. (1993). Using the focus group in assessing training needs Empowering child welfare workers. Child Welfare Journal of Policy Practice and Program.
d Oleire-Oltmanns S. Marzolff I. Peter K. D. & Ries J. B. (2012). "Unmanned Aerial Vehicle (UAV) for monitoring soil erosion in Morocco." Remote Sensing 4(11) 3390-3416.
Eisenbei H. & Zrich E. T. H. (2009). UAV photogrammetry ETH. Ellenberg A. Branco L. Krick A. Bartoli I. & Kontsos A. (2014). "Use of Unmanned
Aerial Vehicle for Quantitative Infrastructure Evaluation." Journal of Infrastructure Systems 21(3) 04014054 Eschmann C. Kuo C. Kuo C. & Boller C. (2012). "Unmanned aircraft systems for remote building inspection and monitoring." Proceedings of the sixth European workshop on structural health monitoring.
151

FAA (2017). Aeronautical Information Manual Official Guide to Basic Flight Information and ATC Procedures. U.S. Department of Transportation. https //www.faa.gov/air_traffic/publications/media/AIM_Basic_dtd_10-12-17.pdf.
Fred Judson (2013) "The Ohio Department of Transportation and Unmanned Aircraft Systems." LiDAR Magazine Vol.3 No.5
Frierson T. (2013). Use of Unmanned Aerial Vehicles for AHTD Applications "Studying Visual Aids to Assist in Corridor Analysis." Retrieved from Arkansas State Highway and Transportation Department
GDOT. (2013). GDOT Offices and Divisions. Retrieved from Georgia Department of Transportation.
Gheisari M. & Esmaeili B. (2016). Unmanned Aerial Systems (UAS) for Construction Safety Applications. Paper presented at the Construction Research Congress 2016.
Gheisari M. Karan E. P. Christmann H. C. Irizarry J. & Johnson E. N. (2015). Investigating Unmanned Aerial System (UAS) Application Requirements within a Department of Transportation. Paper presented at the Transportation Research Board 94th Annual Meeting.
Gillins M. N. Gillins D. T. & Parrish C. (2016). Cost-Effective Bridge Safety Inspections Using Unmanned Aircraft Systems (UAS). Paper presented at the Geotechnical and Structural Engineering Congress 2016.
Gu Y. (2009). Evaluation of Remote Sensing Aerial Systems in Existing Transportation Practices. Retrieved from Virginia State Department of Transportation
152

Gucunski N. Kee S.-H. La H. M. Basily B. & Maher A. (2015). "Delamination and concrete quality assessment of concrete bridge decks using a fully autonomous RABIT platform." Structural Monitoring and Maintenance 2(1) 19-34.
Guerrero J. A. & Bestaoui Y. (2013). "UAV path planning for structure inspection in windy environments." Journal of Intelligent & Robotic Systems 69(1-4) 297-311.
Hallermann N. & Morgenthal G. (2014). Visual inspection strategies for large bridges using Unmanned Aerial Vehicles (UAV). Paper presented at the 7th International Conference on Bridge Maintenance Safety and Management IABMAS.
Hardin P. J. & Jackson M. W. (2005). "An unmanned aerial vehicle for rangeland photography." Rangeland Ecology & Management 58(4) 439-442.
Hart W. S. & Gharaibeh N. G. (2011). Use of micro unmanned aerial vehicles in roadside condition surveys. Paper presented at the Transportation and Development Institute Congress 2011 Integrated Transportation and Development for a Better Tomorrow.
Higuchi K. Shimada T. & Rekimoto J. (2011). "Flying sports assistant external visual imagery representation for sports training." Proceedings of the 2nd Augmented Human International Conference.
Hudzietz B. P. & Saripalli S. (2011). An experimental evaluation of 3d terrain mapping with an autonomous helicopter. Paper presented at the Conference on Unmanned Aerial Vehicle in Geomatics.
Hunt R. C. (2016) The Use of Unmanned Aircraft Systems to Increase Safety and Decrease Costs of Transportation Projects and/or Related Tasks. New Hampshire State Department of Transportation
153

Irizarry J. & Costa D. B. (2016). "Exploratory Study of Potential Applications of Unmanned Aerial Systems for Construction Management Tasks." Journal of Management in Engineering 32(3) 05016001
Irizarry J. and Johnson E. (2014) Feasibility Study to Determine the Economic and Operational Benefits of Utilizing Unmanned Aerial Vehicles (UAVs). Final Report to Georgia Department of Transportation Project 12-38 May 06 2014. (http //www.dot.ga.gov/doingbusiness/research/Documents/12-38.pdf)
Irizarry J. Kim S. Johnson N. E. & Lee K. (2017). Potential Unmanned Aerial Systems Based Operations within a Department of Transportation Findings from a Focus Group Study. Paper presented at the 53rd Annual Associate School of Construction (ASC ) International Conference Seattle.
Irizarry J. Gheisari M. & Walker B. N. (2012). "Usability assessment of drone technology as safety inspection tools." Journal of Information Technology in Construction (ITcon) 17 194-212.
Karan E. P. Christmann C. Gheisari M. Irizarry J. & Johnson E. N. (2014). A comprehensive matrix of unmanned aerial systems requirements for potential applications within a department of transportation. Paper presented at the Construction Research Congress.
Karpowicz R. (2014). The Use of Unmanned Aerial Systems for Steep Terrain Investigations. Retrieved from Caltrans
Khan F. Ellenberg A. Mazzotti M. Kontsos A. Moon F. Pradhan A. & Bartoli I. (2015). Investigation on Bridge Assessment Using Unmanned Aerial Systems. Paper presented at the Structures Congress 2015. 154

Kim S. Irizarry J. & Costa D. B. (2016). Potential Factors Influencing the Performance of Unmanned Aerial System (UAS) Integrated Safety Control for Construction Worksites. Paper presented at the Construction Research Congress 2016.
Kitzinger J. (1995). Qualitative research. Introducing focus groups. BMJ British medical journal 311(7000) 299.
Krajnk T. Vonsek V. Fiser D. & Faigl J. (2011). "AR-drone as a platform for robotic research and education." Research and Education in Robotics-EUROBOT 2011 (pp. 172-186) Springer.
Laa H. M. Gucunski N. Kee S.-H. & Nguyen L. (2014). "Visual and Acoustic Data Analysis for the Bridge Deck Inspection Robotic System." Proceedings of the International Symposium on Automation and Robotics in Construction.
Lin Y. Hyyppa J. Rosnell T. Jaakkola A. & Honkavaara E. (2013). "Development of a UAV-MMS-Collaborative Aerial-to-Ground Remote Sensing SystemA Preparatory Field Validation." IEEE Journal of selected topics in applied earth observations and remote sensing 6(4).
McCormack E. D. and Trepanier T. (2008). The use of small unmanned aircraft by the Washington State Department of Transportation. Washington State Department of Transportation
McGuire M. et al. (2016). A Study of How Unmanned Aircraft Systems Can Support the Kansas Department of Transportation s Efforts to Improve Efficiency Safety and Cost Reduction. Kansas State Department of Transportation.
Metni N. & Hamel T. (2007). "A UAV for bridge inspection Visual servoing control law with orientation limits." Automation in Construction 17(1) 3-10. 155

Moller P. S. (2008). CALTRANS Bridge Inspection Aerial Robot Final Report. Retrieved from California Department of Transportation
NCDOT. (2017). Final Report - North Carolina UAS Airspace Integration Exercise. Retrieved from North Carolina Department of Transportation
Ng W. S. & Sharlin E. (2011). Collocated interaction with flying robots. Paper presented at the RO-MAN 2011 IEEE.
Nisser T. & Westin C. (2006). "Human factors challenges in unmanned aerial vehicles (UAVs) A literature review." School of Aviation of the Lund University Ljungbyhed.
Oskouie P. Becerik-Gerber B. & Soibelman L. (2015). "A data quality-driven framework for asset condition assessment using LiDAR and image data." Computing in Civil Engineering 2015 240-248.
Otero L. D. Gagliardo N. Dalli D. Huang W. & Cosentino P. (2015). Proof of Concept for Using Unmanned Aerial Vehicles for High Mast Pole and Bridge Inspections.
Rinaudo F. Chiabrando F. Lingua A. M. & Span A. T. (2012). "Archaeological site monitoring UAV photogrammetry can be an answer." The International archives of the photogrammetry Remote sensing and spatial information sciences 39(B5) 583-588.
Rodriguez-Gonzalvez P. Gonzalez-Aguilera D. Lopez-Jimenez G. & Picon-Cabrera I. (2014). "Image-based modeling of built environment from an unmanned aerial system." Automation in Construction 48 44-52.
156

Rosnell T. & Honkavaara E. (2012). "Point cloud generation from aerial image data acquired by a quadrocopter type micro unmanned aerial vehicle and a digital still camera." Sensors 12(1) 453-480.
Siebert S. & Teizer J. (2014). "Mobile 3D mapping for surveying earthwork projects using an Unmanned Aerial Vehicle (UAV) system." Automation in Construction 41 1-14.
Sim J. (1998). Collecting and analysing qualitative data issues raised by the focus group. Journal of advanced nursing 28(2) 345-352.
Verhoeven G. Doneus M. Briese C. & Vermeulen F. (2012). "Mapping by matching a computer vision-based approach to fast and accurate georeferencing of archaeological aerial photographs." Journal of Archaeological Science 39(7) 2060-2070.
Werner J. (2003). "FDOT Explores the Viability of Using Unmanned Aerial Vehicles (Uavs) for Traffic Surveillance." Newsletter of the ITS Cooperative Deployment Network [online]
Zink J. & Lovelace B. (2015). Unmanned Aerial Vehicle Bridge Inspection Demonstration Project Minnesota Department of Transportation Research Services & Library.
157

12. Appendix
12.1 Appendix aa IRB Approval Letter
158

12.2 Appendix bb Participant Consent Form
159

12.3 Appendix cc Data Collection Sheet
160

161

162

163

164

165

166

167

12.4 Appendix dd Demographic Information Data Collection Form
168

169

170

12.5 Appendix ee Compilation of Various State UAS Laws
State UAS Laws (compiled from Rupprecht Master List of Drone Laws (Organized by State & Country))

State AL AK AZ AR
CA

Law or Regulation No State laws HB 255 (2014) Local laws
SB 1449 (2016)
HB 1349 (2015) HB 1770 (2015) SB 807 (2016) AB 1680 (2016) AB 856 (2015)
Local laws

Comment
City of Oxford ordinance prohibits flying a drone over city-owned property. Puts limits on how law enforcement can use drones in their operations including how and whether they can save images and video captured by drone. In the Southeastern area during an open commercial salmon fishing period UAVs may not be used for any activity related to commercial salmon fishing operations. UAS cannot interfere with police firefighters or manned aircraft. UAS cannot fly within 500 feet horizontally or 250 feet vertically of
any critical facility. Cities and towns in Arizona that contain more than one park must
allow drones in at least one of them. Cities and towns in Arizona are prohibited from creating their own
drone laws. The Arizona State Legislature claims pre-emption for the creation of any regulations concerning drones. Makes it illegal to use a drone to record someone who has a reasonable expectation of privacy. Prohibits the use of UASs to collect information about or photographically or electronically record information about critical infrastructure without consent. Provides immunity for first responders who damage a UAS that was interfering with the first responder while he or she was providing emergency services. Makes it a misdemeanor to interfere with the activities of first responders during an emergency. Prohibits entering the airspace of an individual to capture an image or recording of that individual engaging in a private personal or familial activity without permission. This legislation is a response to the use of UAS by the press in covering celebrities and other public figures. Town of Los Alamitos ordinance creates restrictions on drone flight and activity within the town. City of Yorba Linda ordinance bans drone takeoffs and landings outside of a drone pilot s visual line of sight within 25 feet of another individual excepting the drone pilot or drone pilot s designee and on private property without the consent of the property owner. This city ordinance also prohibits takeoffs and landings within 500 feet of a special event or emergency response without a city-issued temporary use permit and any violation of an FAA temporary flight restriction or notice to airmen. Town of Calabasas ordinance gives local authorities the power to enforce FAA drone-related regulations by making violations of FAA regulations a misdemeanor. This city ordinance also places limits on how close a drone may fly to a school or public event.
171

In addition to the local laws listed above the National Park Service

band the use of drones in all Golden Gate National Parks in the San

Francisco Bay Area.

Requires the Center of Excellence within the Department of Public

Safety to perform a study to identify ways to integrate UAS within local

and State government functions relating to firefighting search and

rescue accident reconstruction crime scene documentation emergency

CO HB 1070 (2017) management and emergencies involving significant property loss injury

or death. This law also creates a pilot program requiring the deployment

of at least one team of UAS operators to a region of the state that has

been designated as a fire hazard where they will be trained on the use of

UAS for the aforementioned functions.

Prohibits Connecticut municipalities from regulating drones but it does

SB 975 (2017)

allow a municipality that is also a water company to enact ordinances that regulate or prohibit the use or operation of UAS over the

CT

municipality s public water supply and land. Prohibits drone use at Connecticut state parks state forests or other

DEEP 23-4-1 lands under the control of the Department of Energy and Environmental

(2017)

Protection unless specifically authorized by the commissioner in a

special use license.

Makes it illegal to fly a drone over events with more than 5 000 people in

attendance including sporting events concerts automobile races

festivals. It also makes it illegal to fly a drone over critical infrastructure

DE

HB 195 (2016)

which includes but is not limited to oil & gas refiners power plants military facilities government buildings and water treatment facilities.

Finally this law prohibits cities and towns in Delaware from creating

their own drone laws by claiming pre-emption for the creation of all such

laws for the Delaware General Assembly.

Pre-empts local regulation of UAS so that only the Florida legislature

can make laws concerning the use of drones in the State but allows local

governments to enact drone ordinances related to nuisances voyeurism

HB 1027 (2017) harassment reckless endangerment property damage or other illegal

acts. This law also prohibits the operation of drones over or near critical

infrastructure in most instances and prohibits the possession or operation

of a weaponized UAS.

Prohibits the use of a drone to capture an image of privately owned

SB 766 (2015) property or the owner tenant or occupant of such property without

consent if a reasonable expectation of privacy exists.

Defines what a drone is and limits the use of drones by law enforcement.

Under this law law enforcement may use a drone if they obtain a

FL

SB 92 (2013) warrant there is a terrorist threat or "swift action" is needed to prevent

loss of life or to search for a missing person. This law also enables

someone harmed by an inappropriate use of drones to pursue civil action.

Town of Bonita Springs ordinance limits the flying of drones at

Community Park in Bonita Springs to times when the fields of the park

are unoccupied. This city ordinance also makes it illegal to fly within

25 feet of people power lines buildings or light fixtures.

Local laws

City of Miami ordinance prohibits the use of drones over or within a half-mile radius of sporting events or large-venue events including but

not limited to Bayfront Park Marlins Ballpark Miami Marine

Stadium Calle Ocho Festival and any other public parks or facilities

during special events. This city ordinance also prohibits drones from

being equipped with any type of detachable cargo or carrying any type

172

of weapon and establishes a requirement that a city permit is required

for certain "drone-related activities."

Town of Defuniak Springs ordinance prohibits drones from being

flown over public or private property without the owner s consent.

This ordinance also requires that commercial drone pilots register with

the town police department before doing any kind of commercial work

using drones in the city.

City of Orlando ordinance places restrictions on flying drones within

500 feet of city-owned parks schools and venues such as the Amway

Center Camping World Stadium and Harry P. Leu Gardens. This

ordinance also places restrictions on the use of drones within 500 feet

of gatherings with more than 1 000 people. A permit is required to fly

in these areas which costs $20 per flight or $150 annually and those

caught in violation of this ordinance will have to pay fines between

$200 and $400.

Pre-empts local governments in the state from creating UAS regulations

HB 481 (2017) after April 1 2017. This law also allows the regulation of the launch or

landing of UAS on public property by the state or local governments.

City of Conyers ordinance prohibits the use of drones within the

boundaries of the city horse park as well as the Cherokee Run Golf

Course.

City of Augusta ordinance prohibits drone operations in populated

GA

areas within the limits of Richmond County without prior

Local laws

authorization from the FAA and the Augusta Georgia Commission. An exception to this prohibition is the existing model aircraft field at

the intersection of Mike Padgett Highway and Horseshoe Road as

well as any other model aircraft field later approved by the Augusta

Georgia Commission.

Cherokee County ordinance establishes that drones can only be flown

in areas specifically designated for them.

Creates a chief operating officer position for the Hawaii UAS test site.

HI

SB 661 (2015)

This law also establishes an unmanned aerial systems test site advisory board to plan and oversee test site development and appropriates funds to

establish the test site.

SB 1213 (2016)

Prohibits the use of drones for hunting molesting or locating game animals game birds or fur-bearing animals.

ID

Requires warrants for the use of drones by law enforcement establishes

SB 1134 (2013) guidelines for their use by private citizens and provides civil penalties

for damage caused by their improper use.

Loosens regulations around law enforcement s use of UAS during a

disaster or public health emergency and creates regulations for how law

enforcement can obtain and use information gathered from a private

SB 2937 (2014) party s use of drones. This law also requires law enforcement to follow

warrant protocols to compel third parties to share information and if the

information is voluntarily given to police authorities are required to

IL

HB 1652 (2013)

follow the state s law governing drone data retention and disclosure. Prohibits anyone from using a drone to interfere with hunters or fisherman.

Allows drones to be used by law enforcement with a warrant to counter a

terrorist attack to prevent harm to life or to prevent the imminent escape

SB 1587 (2013) of a suspect. If a law enforcement agency uses a drone the agency must

destroy all information gathered by the drone within 30 days but a

supervisor at the law enforcement agency may retain particular

173

information if there is a reasonable suspicion that it contains evidence of

criminal activity.

Village of Schaumburg ordinance prohibits the use of drones within

100 feet of the perimeter of any village property or on any village

Local laws

right-of-way during a special event.

City of Evanston ordinance establishes a moratorium on drone use

until reasonable state and federal regulations are enacted.

Creates new criminal offenses related to the use of drones which

include

The "sex offender unmanned aerial vehicle offense" occurs when a sex

offender uses a UAV to follow contact or capture images or

recordings of someone and when the sex offender is subject to

SB 299 (2017)

conditions that prohibit him or her from doing so. The "public safety remote aerial interference offense" occurs when

someone operates a UAV in a way that is intended to obstruct or

interfere with a public safety official in the course of their duties.

All offenses created by this law are class A misdemeanors. However if

IN

the guilty party has a prior conviction under the same section it becomes

a Level 6 felony.

HB 1013 (2016)

Allows the use of drones to photograph or take video of a traffic crash site.

HB 1246 (2016) Prohibits the use of UAS to scout game during hunting season.

Creates warrant requirements and exceptions for the police use of drones

and real time geo-location tracking devices. This law also creates the

HB 1009 (2014)

crime of "Unlawful Photography and Surveillance on Private Property " making it a Class A misdemeanor defined as knowingly and

intentionally conducting electronic surveillance of the private property of

another without permission.

Illegal for a state agency to use a UAS to enforce traffic laws. This law

IA HB 2289 (2014) requires a warrant or other lawful means to use information obtained via

UAS in a civil or criminal court proceeding.

KS

SB 319 (2016) Local laws

Expands the definition of harassment in the state s Protection from Stalking Act to include certain uses of drones. City of Wichita ordinance bans the use of drones on or near airport property.

Allows commercial airports to prepare UAS facility maps and specifies

that UAS operators cannot operate take off or land in certain areas

designated by an airport s map. This law also prohibits the operation of

UAS in a reckless manner defined as a manner that creates a serious risk

KY

HB 540 (2017)

of physical injury or damage to property. Anyone who violates these provisions is guilty of a class A misdemeanor or a class D felony if the

violation causes a significant change of course or a serious disruption to

the safe travel of an aircraft. In addition this law specifies that these

provisions do not apply to commercial operators in compliance with

FAA regulations.

SB 69 (2017)

Specifies that only the state may regulate UASs pre-empting local regulation.

Adds intentionally crossing a police cordon using a drone to the crime of

LA

SB 73 (2016)

obstructing an officer. This law also allows law enforcement or fire department personnel to disable UAS in the area if they endanger the

public or an officer s safety.

Prohibits using a drone to conduct surveillance of a school school

HB 19 (2016) premises or correctional facilities and establishes a fine of up to $2 000

and up to six months in jail for violations.

174

HB 335 (2016)

Authorizes the establishment of registration and licensing fees for UAS in the State with a limit of $100.

HB 635 (2016)

Adds UAS use to the crimes of voyeurism and video voyeurism in the state.

SB 141 (2016)

Specifies that surveillance by an unmanned aircraft constitutes criminal trespass under certain circumstances.

SB 183 (2015) Regulates the use of UAS in agricultural commercial operations.

HB 1029 (2014)

Creates the crime of unlawful use of an unmanned aircraft system defined as the intentional use of a drone to conduct surveillance of a targeted facility without the owner s prior written consent. This crime is punishable by a fine of up to $500 and imprisonment for six months. A

second offense can be punished with a fine up to $1 000 and one year of

imprisonment.

Requires law enforcement agencies to receive approval before adopting

ME

LD 25 (2015)

the use of drones sets out standards for UAS operation by law enforcement and requires that law enforcement secure a warrant to use

UAS for criminal investigations.

Pre-empts county and municipal authority and specifies that only the MD SB 370 (2015) state can enact laws to prohibit restrict or regulate the testing or
operation of unmanned aircraft systems.

City of Chicopee ordinance states that a drone and/or aircraft shall

only take off and land on private property owned by the operator or

where written permission is granted by the landowner. Said written

permission shall include the name and signature of the land owner the

address of the property and the permissible dates and hours of

MA

No State laws

operations. There are a number of other rules for hobbyist (non-Part 107) operators.

City of Boston policy states that drones may be flown recreationally in city parks so long as FAA policies and safe-flight guidelines are followed.

Town of Holyoke ordinance makes it illegal to fly UAS over privately-

owned or city-owned property without consent.

Prohibits local governments from regulating UASs except when the regulated drone belongs to the locality.

Specifically allows commercial drone operation in the state if the operator is authorized by the FAA to operate commercially and permits hobby operation so long as the operator complies with federal law.

SB 992 (2016)

Prohibits using a drone in a way that interferes with emergency personnel and prohibits the use of a drone to harass an individual to

MI

violate a restraining order or to capture images in a way that invades an individual s reasonable expectation of privacy.

Prohibits sex offenders from using a drone to follow contact or

photograph a person that they are prohibited from contacting.

Anyone who uses a drone in a manner prohibited by this law is guilty of a misdemeanor.

SB 54 (2015)

Prohibits using UASs to interfere with or harass an individual who is hunting.

Local laws

Town of West Bloomfield ordinance establishes all town parks as no-fly zones.

MN

SF 550 (2017) Local laws

Appropriates $348 000 to assess UAS use in natural resource monitoring of moose populations and changes in ecosystems.
Anoka County ordinance requires drone operators to secure a special use permit from the parks department to fly a drone over county parks.

175

Town of St. Bonifacius ordinance bans drones in all city public

airspace.

MS

SB 2022 (2015)

Establishes that using a drone to commit "peeping tom" activities is a felony.

MO

No State laws

Prohibits using UAS to interfere with wildfire suppression efforts.

Anyone who violates this prohibition is liable for the amount of money

HB 644 (2017) equivalent to the costs of their interference. This law also prohibits local

governments from enacting an ordinance addressing UAS use in relation

MT

to a wildfire.

Limits when information gained from UAS use may be admitted as

SB 196 (2013)

evidence in any prosecution or proceeding within the State as only information that was obtained with a search warrant or through a

judicially recognized exception to search warrants.

NE

No State laws

Prohibits the weaponization of UAS and UAS use within a certain

distance of critical facilities and airports without permission. This law

NV AB 239 (2015) also specifies restrictions on UAS use by law enforcement and public

agencies and requires the creation of a registry of all UASs operated by

public agencies in the State.

NH

SB 222 (2015) Prohibits UAS use for hunting fishing or trapping.

Allows UAS operations that are consistent with federal law.

Specifies that UAS owners or operators of critical infrastructure may

apply to the FAA to prohibit or restrict UAS operation near the critical

infrastructure.

Establishes that operating a UAS in a manner that endangers the life or

property of another is a disorderly persons offense.

Establishes that it is a fourth-degree crime if a person "knowingly or

intentionally creates or maintains a condition which endangers the

SB 3370 (2017)

safety or security of a correctional facility by operating an unmanned

aircraft system on the premises of or in close proximity to that

facility."

NJ

Makes it a criminal offense to operate a UAS in a way that interferes

with a first responder.

Defines operating a UAS under the influence of drugs or with a BAC

of .08 percent as a disorderly persons offense.

Pre-empts local governments from regulating UAS in any way that is

inconsistent with this law.

Ramapo Indian Hills ordinance prohibits the use of drones on or above

school grounds.

Local laws

Bernards Township ordinance prohibits the use of drones in or over any park or recreation facility.

Chatham Township ordinance prohibits the use of drones in public

airspace under 400 feet.

NM

SB 556 (2013) Prohibits the use of drones for unwanted surveillance

New York City restriction declares that drones are illegal to fly in New

York City and advises anyone who sees a drone being flown to call

911. This restriction does not seem to be an actual law passed by the

NY

No State laws

city but a policy that the city has adopted. City of Syracuse ordinance bans the use of drones by city officials

until adequate federal and state laws are passed regarding the

government use of drones in a manner that protects citizens First and

Fourth Amendment rights.

176

Prohibits UAS operation near a correctional facility excluding certain

HB 128 (2017) people operating in an official capacity or with written consent from the

warden.

Allows UAS use for emergency management activities including

incident command area reconnaissance search and rescue preliminary

HB 337 (2017)

damage assessment hazard risk management and floodplain mapping. This law also makes other changes to align the state law with federal law

and exempts model aircraft from UAS training and permitting

requirements.

Expands the authority of the state s chief information officer to approve

SB 446 (2015) the purchase and operation of UAS by the state and modifies the state

regulation of UAS to conform to FAA guidelines.

Commercial drone pilots operating in the State of North Carolina must

Commercial UAS/drone operators operating under 14 CFR Part 107 or

a 333 Exemption within North Carolina are required to have a valid

NC UAS Commercial Operators Permit.

NC

Commercial operators must take and pass NCDOT s UAS Knowledge Test and then apply for a state permit.

To obtain a permit operators must provide the state proof of their

remote pilot certificate or other authorization to conduct commercial

UAS operations from the FAA (see Federal above).

SB 744 (2014) Permitted operators agree to these terms & conditions.

Recreational drone pilots flying in North Carolina are not required to

obtain a license or permit from the state s Division of Aviation.

However recreational users are still subject to NC UAS rules and

regulations.

Government/public-use drone pilots operating in the State of North

Carolina must

Take and pass NCDOT s UAS Knowledge Test and then apply for a

State permit.

Agree to these terms & conditions

Town of Chapel Hill ordinance allows local authorities to enforce

Local laws

existing FAA drone regulations.

City of Kannapolis ordinance bans the use of drones in city parks.

ND

HB 1328 (2015)

Provides limitations for the use of UAS for surveillance and prohibits arming a UAS with lethal weapons.

Creates the aerospace and aviation technology committee. One of the

HB 292 (2014) committee s duties is to research and develop aviation technology

including unmanned aerial vehicles.

OH

City of Cleveland ordinance authorizes city police to enforce FAA

Local laws

laws with regards to drones. City of Celina ordinance bans drones in airspace over city-owned

property including parks.

OK

HB 2559 (2016)

Prohibits the operation of UAS within 400 feet of any critical infrastructure facility.

Modifies the law prohibiting UAS weaponization making it a class C

felony to fire a bullet or projectile from a weaponized UAS.

HB 3047 (2017) Allows law enforcement to use UAS to reconstruct an accident scene.

OR

Prohibits the use of UAS over private property in a manner that

intentionally knowingly or recklessly harasses or annoys the owner or

occupant of the property.

HB 4066 (2016)

Modifies definitions related to UAS and makes it a class A misdemeanor to operate a weaponized UAS and regulates the use of drones by public

177

bodies including requiring policies and procedures for the retention of

data. This law also prohibits the use of UAS near critical infrastructure

including correctional facilities.

SB 5702 (2016) Specifies the fees for the registration of public UAS.

Allows a law enforcement agency to operate a drone if it has a warrant

and for enumerated exceptions including for training purposes.

Requires that a drone operated by a public body be registered with the

Oregon Department of Aviation (DOA) which shall keep a registry of

drones operated by public bodies.

Creates new crimes and civil penalties for mounting weapons on

HB 2710 (2013)

drones and interfering with or gaining unauthorized access to public drones.

Allows that under certain conditions a landowner can bring an action

against someone flying a drone lower than 400 feet over their property.

Requires that the DOA must report to legislative committees on the

status of federal regulations and whether UAV s operated by private

parties should be registered in a manner similar to the requirement for

other aircraft.

PA

No State laws

Town of Lower Merion ordinance bans drones in all town parks. City of Pittsburgh ordinance bans drones in city parks or playgrounds.

Gives exclusive regulatory authority over UAS use to the State of Rhode

RI HB 7511 (2016) Island and the Rhode Island Airport Corporation subject to federal law

and pre-empts local governments from creating their own UAS laws.

SC

No State laws

SB 22 (2017)

Exempts UAS aircraft that weigh less than 55 pounds from aircraft registration requirements.

Requires that UAS operation comply with all applicable FAA

requirements.

Prohibits operation of drones over the grounds of correctional and

military facilities making such operation a class 1 misdemeanor. If a

drone is used to deliver contraband or drugs to a correctional facility

SD

SB 80 (2017)

the operator is guilty of a Class 6 felony.

Modifies the crime of unlawful surveillance to include intentional use

of a drone to observe photograph or record someone in a private

place with a reasonable expectation of privacy and landing a drone on

the property of an individual without that person s consent. Unlawful

surveillance is a Class 1 misdemeanor.

Local Laws

City of Aberdeen ordinance permits drone operations in city airspace for hobby or recreational purposes only.

Makes it a crime to fly a drone within 250 feet of a critical infrastructure

SB 2106 (2016) facility for the purpose of conducting surveillance or gathering

information about the facility.

Clarifies that it is permissible for a person to use a UAS on behalf of

HB 2376 (2016) either a public or private institution of higher education rather than just

public institutions.

TN

HB 153 (2015)

Prohibits using a drone to capture an image over certain open-air events and fireworks displays.

Makes it a Class C misdemeanor for any private entity to use a drone to

SB 1777 (2014) conduct video surveillance of a person who is hunting or fishing without

their consent.

Makes it a Class C misdemeanor for a person to use a UAS to

SB 1892 (2014) intentionally conduct surveillance of an individual or their property. This

law also makes it a crime to possess those images (Class C

178

misdemeanor) or distribute and otherwise use them (Class B

misdemeanor).

Enables law enforcement to use drones in compliance with a search

warrant to counter a high-risk terrorist attack and if swift action is

SB 796 (2013) needed to prevent imminent danger to life. Evidence obtained in

violation of this law is not admissible in State criminal prosecutions and

those wronged by such evidence can seek civil remedy.

HB 217 (2017)

Prohibits a person from intentionally knowingly or recklessly chasing actively disturbing or harming livestock through the use of UAS.

Pre-empts local regulation of UAS and exempts UAS from aircraft

registration in the state.

Addresses UAS use by law enforcement allowing use for purposes

unrelated to a criminal investigation.

Requires law enforcement create an official record of UAS use that

provides information regarding the use of the drone and any data

acquired.

SB 111 (2017)

Makes it a Class B misdemeanor to fly a UAS that carries a weapon or has a weapon attached.

Modifies the offense of criminal trespass to include drones entering

and remaining unlawfully over property with specified intent.

UT

Specifies that a person is not guilty of what would otherwise be a

privacy violation if the person is operating a UAS for legitimate

commercial or education purposes consistent with FAA regulations. It

also modifies the offense of voyeurism a Class B misdemeanor to

include the use of any type of technology including UAS to secretly

record video of a person in certain instances.

Allows law enforcement agencies to use an unmanned aircraft system to

HB 296 (2015) collect data at a testing site and to locate a lost or missing person in an

area in which a person has no reasonable expectation of privacy.

Regulates the use of UAS by state government entities establishing that

SB 167 (2014) a warrant is required for a law enforcement agency to "obtain receive or

use data" derived from UAS use.

SB 196 (2014)

Requires law enforcement to obtain a warrant before using drones in a place where an individual has a reasonable expectation of privacy.

Regulates the use of drones by law enforcement and requires law

VT

SB 155 (2016) enforcement to annually report on the use of drones by the department.

This law also prohibits the weaponization of drones.

HB 2350 (2017)

Makes it a Class 1 misdemeanor to use UASs to trespass upon the property of another for the purpose of peeping or spying.

Specifies that the fire chief or other officer in charge of a fire department

SB 873 (2017) has authority to maintain order at an emergency incident including the

immediate airspace.

VA HB 412 (2016) Prohibits UAS regulation by local governments.

HB 2125 (2015)

Requires that a law enforcement agency obtain a warrant before using a drone for any purpose except in limited circumstances.

Prohibits drone use by any state agencies "having jurisdiction over

HB 2012 (2013) criminal law enforcement or regulatory violations" or units of local law

enforcement until July 1 2015.

City of Bellevue Parks & Recreation Department policy declares that

drones are not permitted in Bellevue parks except at Marymoor Park

WA

No State laws

Airfield and 60 Acres Park.

City of Seattle ordinance prohibits drones and other remote-controlled

aircraft in parks.

179

Pierce County ordinance places limits on the use of drones by

government agencies.

WV HB 2515 (2015) Prohibits UAS use for hunting.

SB 338 (2016) Prohibits using a drone to interfere with hunting fishing or trapping.

AB 670 (2016) Prohibits UAS operation over correctional facilities.

Town of Greenfield Prohibits persons from launching or landing a

drone outside of their visual line of sight within one-hundred (100)

feet of any person except the operator and assistant operator within

five hundred (500) feet of any festival event picnic protest or public

assembly of more than one-hundred (100) people in a manner so as to

endanger the safety of any person or property within five hundred

(500) feet of any emergency vehicle which is operating its emergency

lights or siren to any active police fire or emergency response

incident to schools that are in session and jails.

WI

City of Hudson Common Council prevents the use of a drone with the

Local laws

intent to photograph record or observe someone in a place where they have a reasonable expectation of privacy like in their backyard or their

residence and imposes a fine of $200 for violations. Though the State

of Wisconsin already regulates use of drones in this manner this city

ordinance makes it easier for local law enforcement to enforce.

City of Chetek Limits the altitude of drone flights near Chetek

Municipal Airport.

Outagamie County Prohibits drone operation on airport grounds.

City of Green Bay Prohibits drone flight below 400 feet within

specified boundaries of special events including Green Bay Packer

games at Lambeau Field.

Requires the Wyoming Aeronautics Commission to develop rules

regulating where unmanned aircraft can take off and land. The

commission is also permitted to develop reasonable rules regulating the

operation of unmanned aircraft through coordination with the unmanned

WY SF 170 (2017) aircraft industry and local governments. This law also specifies that the

commission does not have the power to regulate unmanned aircraft

operation in navigable airspace and makes it unlawful to land an

unmanned aircraft on the property of another person but operators can

pilot an unmanned aircraft over their own property.

180

12.6 Appendix ff Georgia State UAS Laws
The following texts are excerpts from the Georgia State UAS laws.
House Bill 481 (Kevin Tanner 2017) HB 481 regulates the operation of UASs on public property by State or local governments among other resolutions. Chapter 1 of Title 6 of the Official Code of Georgia Annotated is provided below.
6-1-4 of Official Code of Georgia Annotated
(a) (1) As used in this Code section the term unmanned aircraft system means a powered aerial vehicle that
Does not carry a human operator and is operated without the possibility of direct human intervention from within or on the aircraft
Uses aerodynamic forces to provide vehicle lift
Can fly autonomously or be piloted remotely
Can be expendable or recoverable.
(2) Such term shall not include a satellite.
(b) Any ordinance resolution regulation or policy of any county municipality or other political subdivision of Georgia State regulating the testing or operation of unmanned aircraft systems shall be deemed preempted and shall be null void and of no force and effect provided however that a county municipality or other political subdivision of this state may
(1) Enforce any ordinance that was adopted on or before April 1 2017
(2) Adopt an ordinance that enforces FAA restrictions or provides for or prohibits the launch or intentional landing of an UAS from or on its public property except with respect to the operation of an UAS for commercial purposes.
181

(c) The State through agency or departmental rules and regulations may provide for or prohibit the launch or intentional landing of an unmanned aircraft system from or on its public property.
182

12.7 Appendix gg Sample UAS Operations checklists
183

184

12.8 Appendix hh Focus Group Participants

1 - Billy Cantrell 2 -Robbie Brittain 3 -Jeana Beaudry 4 -Harold D. Mull 5 -Toby M. Hammonds 6 -Bob O Daniels 7 -Darrell Johnson 8 -Jeremy Durrence 9 -Job Walker 10 -Charles Blue 11 -Dana McCrary 12 -Josh Cofer 13 -Lamu Chanthavong 14 -Ariel Hekler 15 -Joseph Robinson 16 -Colette Edmisten 17 -Alan Hood

KCI Tech (Proj. Engineer) GDOT (Construction Proj. Engineer) KCI Tech (Proj. Engineer) GDOT (Director Construction Engineer) GDOT (CPME) State Bridge Inspection Manager Regional Bridge Inspection Engineer (specialist) Regional Bridge Inspection Engineer (specialist) Bridge Inspection Technician Bridge Inspection Supervisor (Specialized Team) Regional Bridge Inspection Engineer (specialist) Bridge Inspection Supervisor (Top-side Team) Rail Management Rail Planner Aviation Project Manager Assist Aviation Project Manager Aviation Safety Data Manager

185