Integrating intersection traffic signal data into a traffic monitoring program

Integrating Intersection Traffic Signal Data into a Traffic Monitoring Program
Final Report
GDOT Research Project TO 02-128; RSCH PROJ 13-10 Georgia Tech Project No. 2006U45
October, 2014 Principal Investigator: Dr. Angshuman Guin The contents of this report reflect the views of the author who is 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.

1.Report No.: FHWA-GA- 2. Government Accession 3. Recipient's Catalog

14-1310

No.:

No.:

4. Title and Subtitle:

5. Report Date: September 2014

Integrating Intersection Traffic Signal Data into

a Traffic Monitoring Program

6. Performing Organization Code:

7. Author(s):

8. Performing Organ. Report No.:

Angshuman Guin, Ph.D.

9. Performing Organization Name and Address: 10. Work Unit No.:

School of Civil and Environmental Engineering

Georgia Institute of Technology

11. Contract or Grant No.:

790 Atlantic Dr.

R.P. 13-10, T.O. 02-128

Atlanta, GA 30332-0355

12. Sponsoring Agency Name and Address: Georgia Department of Transportation Office of Research 15 Kennedy Drive

13. Type of Report and Period Covered:
Final; May 2013-September 2014
14. Sponsoring Agency Code:

Forest Park, GA 30297-2534

15. Supplementary Notes:

16. Abstract:

The objective of this study was to provide the Georgia Department of Transportation with an evaluation of the feasibility of integrating intersection traffic signal data into a traffic monitoring program. Some of the pertinent conclusions from this study are:

Accuracy of 15-minute aggregates of vehicle counts is above 90% at a 95% level of confidence at majority of the study sites

Data quality at some intersections are affected by insufficient offset of the detection zone from the stop bar, leading to queuing of vehicles over the detector and resulting in over-counting of vehicles

Data from intersections using video based detection are of comparable quality as other intersections using inductive loop detectors

For using the intersection signal detector data for traffic monitoring the following cautions need to be exercised:

Discard locations with data spikes i.e. detectors reporting more than 500 vehicles per lane per 15 minutes (threshold can be fine-tuned) at any single data-point

Ensure stability of data connection for completeness of data at remote locations which use wireless data transfers

ii

Data availability does not guarantee data quality. Check plot of historical data for expected traffic patterns before considering use of data in traffic monitoring
Use data from signal detector downstream of midblock location (i.e. use data from two intersections instead of one)
Use data from locations where the inductive loop detectors are physically located upstream of the beginning of turn lanes (if any)
Avoid locations where there is evidence (can be confirmed with site visit during peak period) of vehicles queuing over the detection zone
To improve the usability of intersection signal detector data for traffic monitoring it is recommend that for future installations and maintenance on existing detectors, detection zones are moved further upstream beyond the maximum queue length of a typical peak hour queue and beyond the beginning of the turn lanes. If it is not possible to move detections zones upstream of turn lanes, detectors should be installed on turn lanes as well.
Based on the findings of the study, it is apparent that in majority of the cases, the intersection signal detector data is similar in quality to pneumatic tube count data in terms of both the mean and variability of the errors. However, there are two major concerns. Firstly, the intersection signal detector data does not provide classification data. If this is a major impediment in using the intersection signal detector data in traffic monitoring, this problem should be further investigated to identify possible solutions or workarounds. Secondly, as with any detection technology, the accuracy of the induction loop data feeding into the signal cabinets is dependent on the field deployment characteristics. While there is some variability in the level of error from site to site, in general following the intersection eligibility criteria guidelines should help ensure high quality of data for use in the traffic monitoring program. If the data from a particular site is expected to be used extensively on a long term basis, validation of the data via short term counts is recommended.

17. Key Words:

18. Distribution Statement:

Intersection Signal Data, Highway Performance Monitoring System, Automatic Traffic Recorders, Traffic Monitoring, Induction Loops, Pneumatic Tube, Data Accuracy

19.Security

20.Security

21. Number

Classification (of this Classification (of Pages:

report):

this page):

41

Unclassified

Unclassified

of 22. Price:

Form DOT 1700.7 (8-69)

iii

Table of Contents
List of Figures ............................................................................................................................ vi List of Tables.............................................................................................................................. ix Executive Summary................................................................................................................... 1 Acknowledgments ..................................................................................................................... 3 1.0 Introduction ........................................................................................................................ 4
1.1 Objective........................................................................................................................ 4 2.0 Literature Review ............................................................................................................... 6
2.1 Accuracy of Vehicle Counting Technology .................................................................. 6 2.2 Use of Intersection Signal Data for Vehicle Counts ..................................................... 6 3.0 Methodology....................................................................................................................... 8 3.1 Site Selection ................................................................................................................. 8 3.2 Data Connection .......................................................................................................... 10 3.3 Data Collection ............................................................................................................ 10
3.3.1 Pneumatic Tubes ................................................................................................... 10 3.3.2 Video Processing ................................................................................................... 10 3.4 Analysis steps .............................................................................................................. 12 3.4.1 Validation of Baseline data (Pneumatic Tube data) .............................................. 12 3.4.2 Comparison of Signal Detector Data to Baseline Data ......................................... 12 3.5 Pilot Data Collection Study ......................................................................................... 13 4.0 Results and Discussion ..................................................................................................... 14 4.1 Quality Checks on Baseline Data ................................................................................ 14 4.2 Effect of Presence of Turn Lanes ................................................................................ 18 4.3 Effect of Lane Changes ............................................................................................... 21 4.4 Effect of Presence of Other Intersections .................................................................... 24 4.5 Effect of Distance between Intersections .................................................................... 24 4.6 Effect of Volume ......................................................................................................... 26 4.6.1 Intersections without Queuing............................................................................... 26 4.6.2 Intersections with Queuing.................................................................................... 30 4.6.3 Low Volume Road ................................................................................................ 32 4.7 Effect of Detection Technology .................................................................................. 34 4.8 Summary of Results .................................................................................................... 38 5.0 Conclusions and Recommendations................................................................................. 43
iv

5.1 Findings ....................................................................................................................... 43 5.2 Intersection Eligibility Criteria.................................................................................... 43 5.3 Ideal Intersection for Data Use.................................................................................... 45 5.4 Recommendations ....................................................................................................... 46
5.4.1 Installation Considerations .................................................................................... 46 5.4.2 Communication ..................................................................................................... 46 5.4.3 Future Research ..................................................................................................... 47 6.0 References ........................................................................................................................ 49 7.0 Appendix ......................................................................................................................... 7-0
v

List of Figures
Figure 1: Typical Placement of Offset Loops (Induction Loop Detectors) on Pavement (SR 92 @ Woodstock Road). Image Courtesy: Google Street View......................................................... 9
Figure 2: Typical Placement of Detectors on Pavement. Image Courtesy: Google Maps............ 9
Figure 3: Traffic counting application screen shots: (a) lane setup, (b) buttons with the current count numbers showing. (c) Data collector collecting vehicle count data using application...... 11
Figure 4: Recorded View of Traffic on SR 92 and Woodstock Road from Video that was Post Processed using the Tablet Application ...................................................................................... 12
Figure 5: Time Series Plots for Manual Counts and Pneumatic Tube Counts at Intersection of SR 92 WB and King Road for Lane 1 (Top) and Lane 2 (Bottom) ............................................ 15
Figure 6: Time Series Plots for Manual Counts and Pneumatic Tube Counts at Intersection of SR 92 WB and King Road for Lane 2 (Top) and Aggregates over All Lanes (Bottom) ............ 16
Figure 7: Time Series Plot for Pneumatic Tube Counts Upstream of Intersection and Signal Detector Counts at Intersection of SR 140 EB and River Exchange Parkway for Aggregates over All Lanes ............................................................................................................................. 17
Figure 8: Y-Y Plots for Manual Counts Vs Tube Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 1 (Top-Left), Lane 2 (Top-Right), Lane 3 (Bottom-Left) and Aggregates over All Lanes (Bottom Right)................................................................................. 18
Figure 9: Layout of Test Site Location at Intersection of SR 140 and Steeple Chase Dr; Top: Site Overview; Bottom: Close-up View of Pneumatic Tube Count Site (Image Courtesy: Google Maps) .............................................................................................................................. 19
Figure 10: Layout of Test Site Location at Intersection of SR 141 and Old Alabama Road. Left: Site Overview; Right: Detailed View of Turn Lanes (Image Courtesy: Google Maps) ............. 20
Figure 11: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 EB and Steeple Chase Drive for Lane 1................................................. 20
Figure 12: Time Series Plots for Pneumatic Tube Counts and Signal Detectors Counts Aggregated over All Lanes at Intersection of SR 140 NB and Old Alabama Road ................... 21
Figure 13: Layout of Test Site Location at Intersection of SR 92 and Dials Drive (Image Courtesy: Google Maps) ............................................................................................................. 22
Figure 14: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Dr for Lane 1...................................................................... 22
Figure 15: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Dr for Lane 2...................................................................... 23
Figure 16: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Dr for Aggregates over All Lanes...................................... 23
Figure 17: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 EB and Steeple Chase Dr for Lane 2...................................................... 24
vi

Figure 18: Layout of Test Site Location at Intersection of SR 92 and Roswell Crossing (Image Courtesy: Google Maps) ............................................................................................................. 25
Figure 19: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 3...................................................... 25
Figure 20: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Aggregates over All Lanes...................... 26
Figure 21: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 2...................................................... 27
Figure 22: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Aggregates over All Lanes............................................................................................................................................ 27
Figure 23: Layout of Test Site Location at Intersection of SR 92 and Roswell Crossing (Image Courtesy: Google Maps) ............................................................................................................. 28
Figure 24: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 EB and River Exchange Parkway for Aggregates over All Lanes ............................................................................................................................. 29
Figure 25: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Aggregates over All Lanes ............................................................................................................................. 29
Figure 26: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Lane 1 .................................................................. 30
Figure 27: Time Series Plot for Manual Counts, Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Aggregates over All Lanes .................................. 31
Figure 28: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Aggregates over All Lanes................................. 31
Figure 29: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Aggregates over All Lanes with Count Ceiling per lane at 350 vehicles ..................................................................................................................... 32
Figure 30: Layout of Test Site Location at Intersection of SR 314 and Banks Road (Image Courtesy: Google Maps) ............................................................................................................. 33
Figure 31: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 NB and Banks Rd for Lane 1 ................................................................. 33
Figure 32: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 NB and Banks Rd for Lane 2 ................................................................. 34
Figure 33: Layout of Test Site Location at Intersection of SR 237 and SR 141 (Image Courtesy: Google Maps) .............................................................................................................................. 35
Figure 34: Layout of Detection Zones for Tablet based Manual Counts on video at Intersection of SR 237 and SR 141 (Image Courtesy: Google Maps) ............................................................ 35
Figure 35: Time Series Plot for Manual Counts and Signal Detector Counts at Intersection of SR 237 and SR 141 for Lane 1.......................................................................................................... 36
vii

Figure 36: Time Series Plot for Manual Counts and Signal Detector Counts at Intersection of SR 237 and SR 141for Lane 2........................................................................................................... 36 Figure 37: Time Series Plot for Manual Counts and Signal Detector Counts at Intersection of SR 237 and SR 141 for Lane 3.......................................................................................................... 37 Figure 38: Time Series Plot for Manual Counts and Signal Detector Counts at Intersection of SR 237 and SR 141 for 3 Through-Lanes and All 5 Lanes Including Left Turn Lanes ................... 37 Figure 39: Detection Layout Schematic ...................................................................................... 43 Figure 40: Example of Traffic Counts with and without Data Spikes ........................................ 44 Figure 41: Example of Time Series Plot with well Defined Traffic Volume Pattern ................. 45 Figure 42: Proposed Detection Zone Placement Upstream of Beginning of Turn Lanes (Image Courtesy: Google Maps) ............................................................................................................. 46
viii

List of Tables
Table 1: Final Site List .................................................................................................................. 8 Table 2: Summary of Percentage Errors ..................................................................................... 39
ix

Executive Summary
Many signalized intersections operate under some form of actuated control, in that the intersection approaches have some type of sensor. The most common sensor is an inductive loop which can be configured to provide traffic counts. The Office of Traffic Operations is currently installing traffic signal detectors (inductive loops) in the pavement of various high volume corridors in the Atlanta vicinity. The traffic signal detectors are part of the Regional Traffic Operations Program (RTOP), a multi-jurisdictional, signal timing program with the goal of improving traffic flow and reducing vehicle emissions through improved signal timing. This research project determines if and how the traffic volume data from the inductive loops can be appropriately utilized in the Office of Transportation Data's (OTD) traffic monitoring program.
The objective of this study is to provide the Georgia Department of Transportation with an evaluation of the feasibility of integrating intersection traffic signal data into a traffic monitoring program. Specifically, this project evaluates whether the traffic signal detector data is of comparable quality when compared to portable traffic count data under different traffic flow conditions and different intersection geometries. This study also develops a set of eligibility criteria for identifying intersections that can be used for long term performance monitoring data collection.
The first step in the data quality analysis was to ensure that the pneumatic tube data is of acceptable quality and can be used as control data. A comparison of pneumatic tube count data to manual count data from videos provided evidence that the accuracy of the lane-by-lane 15minute pneumatic counts was between 91% - 98% at a 95% level of confidence, depending on the sites. Occurrences of abnormally high values in some of the data from the signal controllers at some sites investigated in the pilot study, affected the choice of locations to be studied in the second phase. In the second phase of data collection, pneumatic tubes were deployed at multiple locations that varied by intersection geometries, spacing of intersections, traffic volumes, presence of queues etc. Locations with potential of heavy queue formation were oversampled in the study to further investigate the data spiking issues. The pneumatic tube data was supplemented with video based manual data collection for several intersections, including one that used video detection systems for vehicle detection instead of inductive loops. Visual comparisons of the time series plots of the data was followed by computation of the mean percentage errors, mean absolute percentage errors, root mean square error and the 95% bounds of the mean percentage errors.
Some of the pertinent conclusions from this study are:
Accuracy of 15-minute vehicle counts aggregated over all lanes is above 90% at a 95% level of confidence at majority of the study sites
Data quality at some intersections are affected by insufficient offset of the detection zone from the stop bar, leading to queuing of vehicles over the detector and resulting in over-counting of vehicles
Data from intersections using video based detection are of comparable quality as other intersections using inductive loop detectors
For using the intersection signal detector data for traffic monitoring the following cautions need to be exercised:
Discard locations with data spikes i.e. detectors reporting more than 500 vehicles per lane per 15 minutes (threshold can be fine-tuned) at any single data-point
1

Ensure stability of data connection for completeness of data at remote locations which use wireless data transfers
Data availability does not guarantee data quality. Check plot of historical data for expected traffic patterns before considering use of data in traffic monitoring
Use data from signal detector downstream of midblock location (i.e. use data from two intersections instead of one)
Use data from locations where the inductive loop detectors are physically located upstream of the beginning of turn lanes (if any)
Avoid locations where there is evidence (can be confirmed with site visit during peak period) of vehicles queuing over the detection zone
To improve the usability of intersection signal detector data for traffic monitoring it is recommend that for future installations and maintenance on existing detectors, detection zones are moved further upstream beyond the maximum queue length of a typical peak hour queue and beyond the beginning of the turn lanes. If it is not possible to move detections zones upstream of turn lanes, detectors should be installed on turn lanes as well. Based on the findings of the study, it is apparent that in majority of the cases, the intersection signal detector data is similar in quality to pneumatic tube count data in terms of both the mean and variability of the errors. However, there are two major concerns. Firstly, the intersection signal detector data does not provide classification data. If this is a major impediment in using the intersection signal detector data in traffic monitoring, this problem should be further investigated to identify possible solutions or workarounds. Secondly, as with any detection technology, the accuracy of the induction loop data feeding into the signal cabinets is dependent on the field deployment characteristics. While there is some variability in the level of error from site to site, in general following the intersection eligibility criteria guidelines should help ensure high quality of data for use in the traffic monitoring program. If the data from a particular site is expected to be used extensively on a long term basis, validation of the data via short term counts is recommended.
2

Acknowledgments
The authors of this report wish to thank the Georgia Department of Transportation for its support and assistance throughout this effort, in particular the efforts of Kiisa Wiegand, Scott Susten, Ray McEachern, Grant Waldrop, Robert Baker, Paul Tanner and Jane Smith.
3

1.0 Introduction
Many signalized intersections operate under some form of actuated control, in that the intersection approaches have some type of sensor. The most common is an inductive loop which can be configured to provide traffic counts. The Office of Traffic Operations is currently installing traffic signal detectors (inductive loops) in the pavement of various high volume corridors in the Atlanta vicinity. The traffic signal detectors are part of the Regional Traffic Operations Program (RTOP), a multi-jurisdictional, signal timing program with the goal of improving traffic flow and reducing vehicle emissions through improved signal timing. This research project will determine if and how the traffic volume data from the inductive loops can be appropriately utilized in the Office of Transportation Data's (OTD) traffic monitoring program. The existing literature on the subject of intersection traffic signal data pertains to a wide range of topics including: placement and calibration of the embedded loops; how the signal system data should be archived; the appropriate schemas; and the development of scripts to transfer data, process and upload the data. One of the studies (1) focused on equipment design and installation practices, reviewing various loop configurations in terms of ability to provide traffic counts, and selecting instances around the country that are using the exiting loops at traffic signals for counts. However, it appears that the traffic counts used by the different agencies were used only for real-time traffic monitoring; not for federal reporting purposes or as part of a highway performance monitoring program.
Another pilot study (2) demonstrated the feasibility of real-time monitoring of traffic operations at a signalized intersection. Using traffic volumes from existing loops and signal timing information, the control delay (uniform delay plus incremental delay) was estimated on a cycleby-cycle basis using equations provided in the Highway Capacity Manual.
There are several other studies (3, 4) that investigate the use of data from signalized intersections for real-time traffic monitoring. However there is limited amount of research on the accuracy and robustness of the data. The current study aspires to bridge this gap and provide an objective evaluation of the accuracy of the data and feasibility of use as it pertains to federal reporting standards.
1.1 Objective
The overarching objective of this research is to provide the Georgia Department of Transportation with an evaluation of the feasibility of integrating intersection traffic signal data into a traffic monitoring program. Specifically, this project is intended to evaluate whether the traffic signal detector data is of comparable quality when compared to portable traffic count data under different traffic flow conditions and different intersection geometries. The key subobjectives may be summarized as follows:
Conduct a review of the literature on previous efforts at using traffic signal detector data
4

Compare traffic signal detector data and data from portable traffic data collection devices from test sites chosen such that they include a variety of traffic volume levels and geometric and operational characteristics
Develop a methodology for estimating mainline traffic counts (bi-directional) from traffic signal detectors
Make recommendations regarding data quality standards Determine eligibility criteria for intersections to be used for performance monitoring
data collection
5

2.0 Literature Review
There has been an ongoing effort of leveraging the large amount of data continuously generated for real time traffic operations for non-real time uses. Both the traffic operations program and traffic monitoring program typically use the same technology, the most commonly used one being inductive loop based sensors. However, the challenge is typically in the difference in the use of the technology for different purposes. While the exact counts might not be critical in traffic operations, where data is refreshed every few seconds, the error levels become much more critical for traffic monitoring where data is typically aggregated over longer periods of time. Errors in measurement tend to get accumulated over time if there is a bias in a particular direction.
2.1 Accuracy of Vehicle Counting Technology
Inductive loop based vehicle sensors have been in use for a long time in the transportation field and is one of the most well established intrusive detection technologies (Sherry, 2001). Texas Transportation Institute has reported the count accuracy to be 98 percent for properly designed and installed, preformed and standard saw-cut loops (Middlenton, 1999). The Minnesota Department of Transportation, based on field tests, have reported accuracies of 0.1 % - 3% for hourly counts on freeways and 2.8% to 8.6% at intersections (MDOT, 1997; MDOT, 2002). The accuracy of inductive loop varies with environmental conditions and is also dependent on the installation and calibration.
The survey conducted by Sherry (Sherry, 2001) indicated that inductive loops are used in traffic monitoring programs in 50 states and pneumatic tubes are used in 48 states. Pneumatic tubes have been used for vehicle counts since the early 1900s and their accuracy is widely accepted for temporary counts. There is not much recent literature regarding any accuracy studies for pneumatic tube counters. A South Dakota study reported a wide -0.92% to 30% error range for pneumatic tubes (Weber, 1999). McGowen and Sanderson (McGowen, 2011) also reported a wide error range, 6.9% to 27.6%, for 15 minute intervals. However over a two day period the error was in the range of 0.0% to 0.5%.
2.2 Use of Intersection Signal Data for Vehicle Counts
Using traffic operations data for traffic monitoring programs is an attractive idea and naturally, there have been previous efforts towards achieving this. An extensive study was performed by the Institute of Traffic Engineers (ITE) Traffic Engineering Council to study the use of existing loops at signalized intersections for traffic counts. The study examined the use of data from different types of loop configurations including exit loops, queue loops, advance loops and stop line loops for single loops; stop line loops and extension loops for multiple loop configurations; three and four loop configurations; as well as continuous long loops. The study also provided guidance on intersection geometry considerations such as presence of center median or island, loops beyond stop line, road transitions, adjacent lane detection in same and opposite direction of travel, pavement surface, loop crosstalk, use of multiple loops etc. Based on the information provided on the case studies in this report it appeared that as of 2007, the following cities and state DOT have attempted to use signal loop data for vehicle counts:
City of Nashua, NH
City of Fremont, CA
City of Bellevue, WA
6

Minnesota Department of Transportation North Carolina Department of Transportation The survey in the report indicated that these entities used offline polling and retrieval of data at the time the survey was conducted. City of Fremont was the most advanced in this area and they had decided to exclusively use far side system loops for the counts (where far side system loops refer to induction loop detectors configured for vehicle counts and are installed immediately outside the bounds of the intersection, downstream of the intersection). The ITE study provides an appropriate pre-cursor to the current study in that the ITE study provides guidance on best practices for the design and placement of the loops. The current study builds on this effort and investigates the use of data from pre-installed offset (or advance) loops or other technologies used for vehicle detection at intersections (such as video detection systems) in a traffic monitoring program.
7

3.0 Methodology

3.1 Site Selection
An inventory of existing signalized intersection sites in Georgia to which the Office of Traffic Operations has data access was obtained. A list of potential sites to study was based on criteria such as geometric features, area characteristics (urban, suburban, rural), current traffic demands (vehicle volumes, percent trucks, etc.), and location. In addition, an initial snapshot of the signal detector data was obtained for the potential sites and reviewed to ensure that the data does not exhibit any abnormal artifacts.
As the study progressed, the initial list of study sites had to be modified to eliminate the initially identified rural site because of unavailability of data at the site. In addition the case study of intersection with data on all four approaches was eliminated because none of the intersections with loop detectors had detection on all four approaches. Some of the initial site selections had to be modified because of unavailability of appropriate anchor points for the pneumatic tubes near the proposed data collection location. The study sites were selected to cover different geometric feature scenarios and traffic demands. Figure 1 shows the typical location of an offset loop at an intersection. Figure 2 shows the typical setup of the different kinds of detection zones at an intersection on the RTOP corridors. Table 1 lists the final set of intersections along with the corresponding selection criteria, where data were collected for this study.

Table 1: Final Site List

Intersection/Location

Criteria

Characteristics

1

SR 141 at East Jones Bridge Rd

High Volume Intersections

2 SR 141 at Old Alabama Rd

3 SR 92 at Woodstock/King Rd

4 SR 140 at River Exchange Parkway

5 Between SR 140 at Fouts

Distance

Multiple cross streets between

Road and SR 140 at

between

intersections

Steeplechase Drive

intersections

6

Between SR 92 at Woodstock/King Rd and SR 92 at Roswell Crossing

(multiple cross streets vs. low cross streets)

No significant cross streets. Distance between intersections is 900 ft.

7 Between SR 92 at Woodlands Dr and SR 92 at Dials Dr

No significant cross streets. Distance between intersections is 1450 ft.

8 SR 314 between Creek Wood Medium/Low

Trl and Banks Rd

Volume

9 SR 237 at SR 141 (Piedmont VDS Based

8

and Peachtree)

Detection

Detection Zone
Figure 1: Typical Placement of Offset Loops (Induction Loop Detectors) on Pavement (SR 92 @ Woodstock Road). Image Courtesy: Google Street View

Offset Count Detectors

Presence Detectors

Stop Bar

Offset Count Detectors
Beginning of Turn Lane

Figure 2: Typical Placement of Detectors on Pavement. Image Courtesy: Google Maps
9

3.2 Data Connection
To test the feasibility of using the signal detector data for counts, it is essential to ensure the availability of the data. The 15-minute aggregated data from the signal detectors are polled on a regular basis (usually once or twice daily) by a central server and the data is archived in a database. In addition, the central server receives real time data from the signal detectors for real-time operations purposes. This data is available to external applications as an XML data packet upon request from the server. To leverage the higher resolution of the real-time data stream, a script was setup on a remote server (at Georgia Tech) to obtain the real time data and archive it. However, data quality analysis indicated that the completeness of the data in the real-time stream could not be guaranteed. Therefore, the final solution involved a script for obtaining a daily snapshot of the 15-minute aggregated data from the database for analysis/processing and archiving.
3.3 Data Collection
Due to the resource intensive nature of video based data collection, the baseline data was collected using pneumatic tubes similar to a standard portable data collection effort conducted by GDOT for the traffic monitoring program. To ensure that the pneumatic tube data is reliable for use as baseline data for comparing with traffic signal detector data (also referred to as signal detector data or simply signal count data in this document), the research team conducted quality checks on this data using video based manually collected data.
3.3.1 Pneumatic Tubes
Portable count data was collected using pneumatic tube counters. The tubes were laid across the roadway as close as possible to the location of the in-pavement inductive loops detectors of the signal detectors. In several cases it was not possible to match the exact location because of constraints regarding the availability of anchor points for the pneumatic tube equipment. In such cases, the analysis has to take into account that differences in counts between the pneumatic tubes and the inductive loops could arise due to lane changes of the vehicles and the time offset between the arrivals of vehicles at the two detectors. The time-offset is typically minor during medium to light volume conditions but could become a significant factor under high volume conditions that lead to vehicle queuing over the detectors.
A total of six hundred and seventy two hours of directional data over nine study sites around Metro Atlanta was collected for analysis this study.
3.3.2 Video Processing
For validating the pneumatic tube counts, manual vehicle count data was collected. To ensure repeatability and high quality of the data, counts were performed using pre-recorded video at the count locations. The video was processed by data collectors using a tablet application that was developed at Georgia Tech (as part of TO 02-69; RSCH PROJ 10-03). Screenshots of the application interface can be seen in Figure 3. The data was obtained at a per-vehicle record level and was aggregated up to 15 minute bins to match the data from the pneumatic tubes. The video data was also used for validation of the data at intersections that used Video Detection Systems for counts, because the extremely high volume of traffic at these sites and the lack of anchor points made it impossible to collect data using pneumatic tube counters.
About 21 hours of per-lane manual vehicle count data was collected and analyzed for this study.
10

(a)
(b)
(c)
Figure 3: Traffic counting application screen shots: (a) lane setup, (b) buttons with the current count numbers showing. (c) Data collector collecting vehicle count data using application
11

Figure 4: Recorded View of Traffic on SR 92 and Woodstock Road from Video that was Post Processed using the Tablet Application
3.4 Analysis steps
3.4.1 Validation of Baseline data (Pneumatic Tube data) To ensure that the pneumatic tube counter data was of a quality sufficient to serve as a baseline, the pneumatic tube data was validated against manually collected data. To ensure repeatability and high quality of the data, counts were performed using pre-recorded video at the count locations. 3.4.2 Comparison of Signal Detector Data to Baseline Data Essential to the evaluation of the feasibility of integrating intersection traffic signal data into a traffic monitoring program is an evaluation of whether the traffic signal detector data is of a quality comparable to portable traffic count data under different traffic flow conditions and different intersection geometries. This is to ensure that the quality of data from the new system (signal detector) that is supplementing the old system (portable counts) is not inferior to the old system data. In addition, it is necessary to evaluate the accuracy of both the systems to ensure that apart from the comparative differences, the actual errors in the new system are not extremely high. Multiple paired comparisons were performed across the different datasets to check for both these scenarios.
3.4.2.1 Visual Analysis The first step in the comparison was a visual analysis of the data to check for large differences or issues in the underlying dataset. 3.4.2.1.1 Time Series Plots
12

The 15-minute aggregates of counts were plotted as time-series charts with counts on the y-axis and time of day on the x-axis for the entire period of data collection. Both the test data and the baseline data were plotted on the same chart to enable easy comparison of the datasets. The time-series charts helped identify any systematic divergence between the values or characteristic attributes of the divergence, e.g. whether the differences are higher during certain flow conditions, time-of-day etc. 3.4.2.1.2 Y-Y Plots The Y-Y plots between the test data and the baseline data shows the relationship of the magnitude of the error values to the magnitude of the actual values. This provided a timeindependent view of the errors. A green line was drawn in these plots to serve as the guide line for the 45 degree line on which the data-points would lie in case of zero error. In addition two red lines were drawn to identify the 10% error bounds. 3.4.2.2 Percentage Error The following statistics were used to compare the differences between the observed counts and the designated baseline counts.
Mean Percentage Error (MPE) Mean Absolute Percentage Error (MAPE) Root Mean Squared Percentage Error (RMSPE)
For the inductive loop detector based signal detector counts, the pneumatic tube counts served as the baseline. For the pneumatic tube counts as well as the video detection system based signal detector counts, the manual counts served as the baseline. Apart from the above measures of central tendency, the variability in the differences was studied by using the 95th percent confidence intervals on the mean percentage error values.
3.5 Pilot Data Collection Study
A pilot study with three adjacent intersections and three intermediate pneumatic tube data collection sites was performed to test the experimental setup. In this case study, the pilot study results matched up very well with expectations and no modifications were necessary to the experimental setup. However, the pilot data revealed some artifacts in the data under high volume conditions which affected the site selection plan. Additional high-volume sites with possible vehicle queuing over the detection station were chosen to further study any possible issues under high-volume traffic flow conditions. The results of the pilot data collection study as well as the full study are presented in Section 4.0.
13

4.0 Results and Discussion
This section describes the results of the comparative analysis. The different scenarios are analyzed on a case by case basis to observe the biases in the datasets or other issues in the data under the influence of different conditions such as geometric characteristics of intersections or flow conditions of the traffic. The subsections use the visuals, i.e. the time series plots, Y-Y plots etc. for discussion. Only the most relevant figures are presented during the discussions. For comprehensive reference, the entire set of charts for each site is provided in the Appendix, where the charts are grouped by site for ease of access. The MPE, MAPE and RMSPE values are provided as a summary table, along with the summary of the results at the end of this section.
4.1 Quality Checks on Baseline Data
Based on the results of initial comparisons between the signal detector data and the pneumatic tube data obtained in the pilot study, the quality of the portable traffic data collection devices was deemed to be acceptable for purposes of this study. In view of concerns about pneumatic tubes undercounting traffic under slow speed and high volume conditions, a detailed comparison of video based counts collected by the research team and the pneumatic tube counts was performed. The pneumatic tube counter data and the manual count data are plotted as time series in Figure 5 and Figure 6 for the SR 92 at the Woodstock / King Road intersection. Y-Y plots are provided in Figure 8. The time series plots show that the tube count data is slightly higher than the manual counts in lanes 1 and 3, the fastest and slowest lanes respectively, but lower than the manual counts in lane 2 (the middle lane). The Y-Y plots show that while most of the data are within the 10% bounds, there are some points in the lane 3 data that fall outside the bounds. Overall, when aggregated across the lanes the data show a slight positive bias. The mean percentage errors vary within a range of -6% to 8.7% for the individual lane aggregates while the aggregate over all lanes has a mean error of 3%. For the SR 140 at River Exchange Parkway site, the results were better at the individual lane level (Figure 7), but similar at the aggregate over lanes level. The mean percentage error varied in the range of 2.3 to 3.1 for the individual lanes and was 2.7% for the aggregate over lanes in the eastbound direction. A part of these errors can be explained away by the possibility of lane changes between the location of the tube count and the location of the screen-line for the manual count. However, the magnitude of the errors is too high to attribute completely to lane changes and differences in the arrival times at the two detection points. Due to the high degree of confidence in the manual data, it can be concluded that there is a certain amount of error in the pneumatic tube counts. While the error in the aggregates across lanes is 3%, the errors in counts in the individual lanes is likely to be higher (slightly less than the -6% to 8.7% range, if lane changing is taken into account).
14

15

Figure 5: Time Series Plots for Manual Counts and Pneumatic Tube Counts at Intersection of SR 92 WB and King Road for Lane 1 (Top) and Lane 2 (Bottom)

10/29/2013 16:30 10/29/2013 16:45 10/29/2013 17:00 10/29/2013 17:15 10/29/2013 17:30 10/29/2013 17:45 10/29/2013 18:00 10/29/2013 18:15 10/29/2013 18:30 10/29/2013 18:45 10/29/2013 19:00 10/29/2013 19:15 10/29/2013 19:30

0

50

100

150

200

250

Manual WB2

300

400 Tube Count
350

10/29/2013 16:30 10/29/2013 16:45 10/29/2013 17:00 10/29/2013 17:15 10/29/2013 17:30 10/29/2013 17:45 10/29/2013 18:00 10/29/2013 18:15 10/29/2013 18:30 10/29/2013 18:45 10/29/2013 19:00 10/29/2013 19:15 10/29/2013 19:30

0

50

100

150

200

250

WB1+LTL

300

Manual

350

Tube + LTL

400

16

900 800 700 600 500
Tube WB All Ln 400
Manual WB All Ln 300 200 100
0
Figure 6: Time Series Plots for Manual Counts and Pneumatic Tube Counts at Intersection of SR 92 WB and King Road for Lane 2 (Top) and Aggregates over All Lanes (Bottom)

10/29/2013 16:30 10/29/2013 16:45 10/29/2013 17:00 10/29/2013 17:15 10/29/2013 17:30 10/29/2013 17:45 10/29/2013 18:00 10/29/2013 18:15 10/29/2013 18:30 10/29/2013 18:45 10/29/2013 19:00 10/29/2013 19:15 10/29/2013 19:30

10/29/2013 16:30 10/29/2013 16:45 10/29/2013 17:00 10/29/2013 17:15 10/29/2013 17:30 10/29/2013 17:45 10/29/2013 18:00 10/29/2013 18:15 10/29/2013 18:30 10/29/2013 18:45 10/29/2013 19:00 10/29/2013 19:15 10/29/2013 19:30

0

50

100

150

200

250

Manual WB3+RTL

300

350

400 Tube + RTL

900

800

700

600

500

Manual All Ln

400

Tube All Ln

300

200

100

0

7:15

7:30

7:45

8:00

8:15

8:30

8:45

9:00

9:15

Figure 7: Time Series Plot for Pneumatic Tube Counts Upstream of Intersection and Signal Detector Counts at Intersection of SR 140 EB and River Exchange Parkway for Aggregates over All Lanes

17

500

450

400

350

300

250

200

150

100

50

0

500

500

450

450

400

400

350

350

300

300

250

250

200

200

150

150

100

100

50

50

0

0

500

450

400

350

300

250

200

150

100

50

0

500

1000

450

900

400

800

350

700

300

600

250

500

200

400

150

300

100

200

50

100

0

0

1000

900

800

700

600

500

400

300

200

100

0

500

450

400

350

300

250

200

150

100

50

0

Figure 8: Y-Y Plots for Manual Counts Vs Tube Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 1 (Top-Left), Lane 2 (Top-Right), Lane 3 (Bottom-Left) and Aggregates over All Lanes (Bottom Right)

4.2 Effect of Presence of Turn Lanes
The challenge with obtaining traffic counts close to an intersection instead of at the midblock is with capturing the turning vehicles. If the detection zones are sufficiently upstream of the beginning of the turn lanes, all the vehicles should be accounted for. However, if the detection zones are downstream of the beginning of the turn lanes, a certain percentage (up to 100%) of the turning vehicles will be missed from the counts if the detection zones are solely on the through lanes. The percentage of loss depends on the traffic behavior near the intersection, i.e. whether the vehicles change lanes into the turn lanes before the detection zone or after.
The other potential impact of the presence of high turning volumes (usually evidenced by the presence of turn lanes or bays) is the associated high amount of lane changing. If a significant
18

portion of the lane changes occur in the detection zone, it could lead to faulty counts (over- or under-counts depending on the wheel path). The absence of turn lanes did not show a direct relationship with the degree of match between the signal count data with the baseline data. SR 140 at Steeple Chase Drive (Figure 9) was the site with no turn lanes. While the data on the eastbound lanes showed a -5.7% error, the westbound lanes had an error of 0.7%. At the SR141 site at Old Alabama Road (Figure 10), there is a significant amount of traffic on the turn lane, but when the volumes on the through lanes are considered, the error levels are low (-2.6% NB and 0.7% SB). The natural errors in the data steam, for example the under-counts around 1 pm in Figure 11, probably contributed significantly towards the higher error level on the eastbound lanes. At the study sites there is not sufficient evidence of lane changing producing a significant effect on the error levels. However it was observed that a significant amount (close to 100%) of the turning traffic was missed by the intersection signal detectors. In Figure 12, the signal detectors counts aggregated over the through lanes match up very well with the counts on those two lanes from the pneumatic tube counts. However, if the counts from the signal detectors on the two through lanes was used to represent the entire volume on that section of the road, the counts will be short by more than 15%.
Figure 9: Layout of Test Site Location at Intersection of SR 140 and Steeple Chase Dr; Top: Site Overview; Bottom: Close-up View of Pneumatic Tube Count Site (Image Courtesy: Google Maps)
19

Figure 10: Layout of Test Site Location at Intersection of SR 141 and Old Alabama Road. Left: Site Overview; Right: Detailed View of Turn Lanes (Image Courtesy: Google Maps)

400

Tube Count

350

Signal Count SR 92 EB Ln 1 @ Steeplechase

300

US Signal (Fouts)

DS Signal (Champions Green) 250

200

150

100

50

0

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 2/18/2014 10:00 2/18/2014 12:30 2/18/2014 15:00 2/18/2014 17:30 2/18/2014 20:00 2/18/2014 22:30

Figure 11: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 EB and Steeple Chase Drive for Lane 1
20

800 Tube Count FL+ML+SL

700

Tube Count FL + ML

NB Signal Count

600

500

400

300

200

100

0

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 2/18/2014 10:00 2/18/2014 12:30 2/18/2014 15:00 2/18/2014 17:30 2/18/2014 20:00 2/18/2014 22:30

Figure 12: Time Series Plots for Pneumatic Tube Counts and Signal Detectors Counts Aggregated over All Lanes at Intersection of SR 140 NB and Old Alabama Road
4.3 Effect of Lane Changes
While no adverse effect of lane changing was observed at the Old Alabama Road site, some amount of lane changing was observed between the Pneumatic Tube data collection site and the intersection signal detector's detection zone for the study site on SR 92 between Woodlands Parkway and Dials Drive (Figure 13). There is a slight positive bias observed in the Lane 1 counts (Figure 14) while there is a slight negative bias observed in the Lane 2 counts (Figure 15). However, the total counts match quite well (Figure 16). The overall impact of lane changes at this site, where there are no other ingress or egress points between the detection zones is, as expected, minimal.

21

Figure 13: Layout of Test Site Location at Intersection of SR 92 and Dials Drive (Image Courtesy: Google Maps)

400

Tube Count

350

SR 92 EB Ln 1 (Dials)

DS Signal (Wigley)

300

US Signal (Ragsdale)

250

200

150

100

50

0

2/20/2014 0:00 2/20/2014 2:30 2/20/2014 5:00 2/20/2014 7:30 /20/2014 10:00 /20/2014 12:30 /20/2014 15:00 /20/2014 17:30 /20/2014 20:00 /20/2014 22:30

Figure 14: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Dr for Lane 1
22

400

Tube Count

350

SR 92 EB Ln 2 (Dials)

DS Signal (Wigley)

300

US Signal (Ragsdale)

250

200

150

100

50

0

2/20/2014 0:00 2/20/2014 2:30 2/20/2014 5:00 2/20/2014 7:30 /20/2014 10:00 /20/2014 12:30 /20/2014 15:00 /20/2014 17:30 /20/2014 20:00 /20/2014 22:30

Figure 15: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Dr for Lane 2

800

Tube Count

700

Signal Count SR 92 EB at Dials

DS Signal (Wigley)

600

US Signal (Ragsdale)

500

400

300

200

100

0

2/20/2014 0:00 2/20/2014 2:30 2/20/2014 5:00 2/20/2014 7:30 /20/2014 10:00 /20/2014 12:30 /20/2014 15:00 /20/2014 17:30 /20/2014 20:00 /20/2014 22:30

Figure 16: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Dr for Aggregates over All Lanes

23

4.4 Effect of Presence of Other Intersections
There are several ingress and egress points between the Fouts Road and Steeple Chase Drive intersections on SR 92 (Figure 9). The site without any major ingress and egress point between the intersections is SR 92 between Woodlands Parkway and Dials Drive (Figure 13). While the match between the signal detector counts and pneumatic tube counts are very well timesynchronized at the Dials Drive site (Figure 14, Figure 15, and Figure 16), there is some amount of volatility of the differences between the counts at the Steeple Chase Drive site (Figure 11 and Figure 17). The differences in the level of error are clear in the fast lane (7.5% versus 11.1% MAPE at Dial Road and Steeple Chase Drive sites respectively), but not that apparent on the slow lane (3.9% versus 4.1% MAPE).

400 Tube Count

350

Signal Count SR 92 EB Ln 2 @ Steeplechase

300

US Signal (Fouts)

DS Signal (Champions Green) 250

200

150

100

50

0

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 /18/2014 10:00 /18/2014 12:30 /18/2014 15:00 /18/2014 17:30 /18/2014 20:00 /18/2014 22:30

Figure 17: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 EB and Steeple Chase Dr for Lane 2
4.5 Effect of Distance between Intersections
The distance between the intersections at the SR 92 site between Woodstock Road and Roswell Crossing is about 900ft (Figure 18) while the distance between the intersections at the SR 92 site between Woodlands Parkway and Dials Drive is about 1450 ft. Both sites are characterized by the absence of any crossroads between the intersections. The site with higher separation (Figure 14, Figure 15, and Figure 16) was expected to have higher variability of the errors, than the site with smaller separation (Figure 19 and Figure 20). But the observed results did not support this hypothesis. This can be explained to some extent by the presence of a right turning lane at the Roswell Crossing intersection as well as the present of a congested intersection (Woodstock Road) in close proximity, which could have negatively affected the data quality at the Roswell Crossing intersection.
24

Figure 18: Layout of Test Site Location at Intersection of SR 92 and Roswell Crossing (Image Courtesy: Google Maps)

400 Tube Count

350

92 @ RosCr EB3

300

250

200

150

100

50

0

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure 19: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 3
25

1000 900 800 700 600 500 400 300 200 100 0

Tube Count 92 @ RosCr EB All Ln

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure 20: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Aggregates over All Lanes
4.6 Effect of Volume
4.6.1 Intersections without Queuing
4.6.1.1 SR 92 at Roswell Crossing
As explained in section 3.1, there were some anomalies observed at a high volume site during the pilot study. This prompted an oversampling of the high volume sites. The SR 92 intersection at Roswell Crossing was adjacent to the congested Woodstock Road intersection. But this signal had a higher amount of green time on SR 92 and therefore experienced very little queuing in the eastbound direction where it is downstream of the Woodstock Road intersection. On the eastbound approach the intersection signal counts matched up very well with lanes 1 and 2 of the pneumatic tube counts (Figure 21). However, the counts on lane 3 (Figure 19) did not match up too well. It is possible that the right turning traffic from Woodstock Road might have caused the queue on the slow extend up to the detection zone under heavy traffic conditions. While the mean percentage error over all lanes was 4.3% (Figure 20), the mean percentage error on the slow lane was 24.3%. Comparison with the pneumatic tube counts downstream of the intersection (Figure 22) yielded a mean percentage error in the same range (21.7%).

26

400 Tube Count

350

92 @ RosCr EB2

300

250

200

150

100

50

0

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure 21: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 2

1000 900 800 700 600 500 400 300 200 100 0

Tube Count D/S 92 @ RosCr EB All Ln

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure 22: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Aggregates over All Lanes
27

4.6.1.2 SR140 @ River Exchange Parkway The results at another high volume site, SR 140 at River Exchange Parkway (Figure 23), showed some undercounting under high volume conditions in a few instances (Figure 24 and Figure 25). But the overall aggregate mean percentage error was in a range of -9.7% to 4.4%.
Figure 23: Layout of Test Site Location at Intersection of SR 92 and Roswell Crossing (Image Courtesy: Google Maps)
28

900 Tube Count EB All Ln

800

Signal Count EB All Ln

700

600

500

400

300

200

100

0

5/5/2014 12:00 5/5/2014 16:48 5/5/2014 21:36
5/6/2014 2:24 5/6/2014 7:12 5/6/2014 12:00 5/6/2014 16:48 5/6/2014 21:36 5/7/2014 2:24 5/7/2014 7:12 5/7/2014 12:00

Figure 24: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 EB and River Exchange Parkway for Aggregates over All Lanes

900 Tube Count WB All Ln

800

Signal Count WB All Ln

700

600

500

400

300

200

100

0

5/5/2014 12:00 5/5/2014 16:48 5/5/2014 21:36
5/6/2014 2:24 5/6/2014 7:12 5/6/2014 12:00 5/6/2014 16:48 5/6/2014 21:36 5/7/2014 2:24 5/7/2014 7:12 5/7/2014 12:00

Figure 25: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Aggregates over All Lanes
29

4.6.2 Intersections with Queuing
The spikes in the data under congested conditions that can be observed in Figure 26 are observed in all three lanes at this site. Figure 27 shows a three way comparison of manual counts, pneumatic tube counts and signal detector counts and shows that the errors in the signal detector counts here are much larger than the error bounds of the pneumatic tube count data. Figure 28 shows that the errors are compounded when the data is aggregated across lanes.
If a basic filter using a threshold of 350 vehicles per 15-minutes per lane (translating to 1400 vehicles per hour per lane) is used to smooth out the spikes in the data, the data appears to get rid of the high volume spikes as shown in Figure 29. However, the entire afternoon peak periods record volumes significantly higher than the baseline counts. While the extremely high spikes are easy to identify, the dataset also includes over-counts that are too high and are difficult to identify and eliminate. If a per-vehicle-record based data stream is available, it might be possible to design a filter to completely eliminate the erroneous volumes, but it is difficult to eliminate such faulty data from the 15-minute aggregates.
Several other high volumes sites such as SR 92 at River Exchange Parkway were also studied (section 4.3.1) and volumes above 250 vehicles per 15-minute period per lane were observed, but evidence of these spikes in volumes were not seen. It is hypothesized that the erroneous volumes are a result of queuing over the induction loop detectors and compounded by some calibration issues.

1400 1200 1000
800 600 400 200
0

Tube + LTL 92 @ King WB1

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure 26: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Lane 1
30

2500 2000 1500 1000
500 0

Tube WB All Ln Signal WB All Ln Manual WB All Ln

10/29/2013 16:30 10/29/2013 16:45 10/29/2013 17:00 10/29/2013 17:15 10/29/2013 17:30 10/29/2013 17:45 10/29/2013 18:00 10/29/2013 18:15 10/29/2013 18:30 10/29/2013 18:45 10/29/2013 19:00 10/29/2013 19:15 10/29/2013 19:30

Figure 27: Time Series Plot for Manual Counts, Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Aggregates over All Lanes

3000 2500 2000 1500 1000
500 0

Tube Count All Ln 92 @ King WB All Ln

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure 28: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Aggregates over All Lanes
31

3000 2500 2000 1500 1000
500 0

Tube Count All Ln
92 @ King WB All Ln Ceiling 350

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure 29: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Aggregates over All Lanes with Count Ceiling per lane at 350 vehicles
4.6.3 Low Volume Road
SR 314 near Banks Road and Creek Wood Trail was a low volume road study site (Figure 30). This site experienced some connectivity issues leading to data losses in the intersection signal data stream. The available data, matched very well with the pneumatic tube counts. The pneumatic tube counts were matched with the upsteam signal counts as well to investigate if, under these low volume conditions, detection from a single intersection can be used instead of requiring a pair of intersections with available data. While Lane 1 upstream data matched very well (Figure 31), the Lane 2 upstream data match was not good (Figure 32), possibly due to differences caused by turning vehicles at the intersection.

32

Figure 30: Layout of Test Site Location at Intersection of SR 314 and Banks Road (Image Courtesy: Google Maps)
200 Tube Count
180 Signal Count SR314 NB Ln 1 @ Banks Rd
160 Signal Count SR314 NB Ln 1 @ Creek Wood (U/S)
140 120 100
80 60 40 20
0
Figure 31: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 NB and Banks Rd for Lane 1
33

5/13/2014 0:00 5/13/2014 4:48 5/13/2014 9:36 5/13/2014 14:24 5/13/2014 19:12 5/14/2014 0:00 5/14/2014 4:48 5/14/2014 9:36 5/14/2014 14:24 5/14/2014 19:12 5/15/2014 0:00

200

180

Tube Count

160

Signal Count SR314 NB Ln 2 @ Banks Rd

140 Signal Count SR314 NB Ln 2 @ Creek Wood (U/S)
120

100

80

60

40

20

0

5/13/2014 0:00 5/13/2014 4:48 5/13/2014 9:36 5/13/2014 14:24 5/13/2014 19:12 5/14/2014 0:00 5/14/2014 4:48 5/14/2014 9:36 5/14/2014 14:24 5/14/2014 19:12 5/15/2014 0:00

Figure 32: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 NB and Banks Rd for Lane 2
4.7 Effect of Detection Technology
Most of the intersections in the region have induction loop based detection. There are a handful of intersections with Video Detection System based detection. Typically, intrusive technologies such as induction loops have a higher level of accuracy of vehicle count data than video detection systems that are easily affected by external factors such as shadows, precipitation, wind and vehicle induced vibration of the camera. In this study, the accuracy of counts from a video detection system installed at the intersection of SR 237 and SR 141 (Figure 33) was tested against manual counts. The manual counts were extracted using the same video feed that was used by the video detection system. The camera is installed on the signal pole across the intersection. The detection zones are set up past the stop bar, which ensures the vehicles are moving (except in case of a gridlock) when they cross the detection zone, and thereby provides higher accuracy of detection.
The accuracy test had two major limitations:
Given the resource intensive nature of the manual data reduction of the videos that were of low quality (Figure 34), the test was limited to a single approach (SR 141 EB) and a single morning peak period.
The only adverse condition that was tested was the transition in lighting conditions (dawn). The other condition that might have been tested is camera vibration due to heavy vehicle traffic, but any movement of the camera was not noticeable in the observed video. Effects of precipitation, high wind, shadows etc. were not tested in this study.
34

As seen in Figure 35, Figure 36, and Figure 37 the errors were low except in two time periods on the middle lane. The video detection system was configured to detect vehicle counts on only three of the lanes. The left turning vehicles were not counted and are therefore unaccounted for (Figure 37). Discounting the left turn lanes, the mean percentage errors on the other three lanes were in the range of 2.6 to 11.8%.
Figure 33: Layout of Test Site Location at Intersection of SR 237 and SR 141 (Image Courtesy: Google Maps)
Figure 34: Layout of Detection Zones for Tablet based Manual Counts on video at Intersection of SR 237 and SR 141 (Image Courtesy: Google Maps)
35

160 Signal Count Ln 1
140 Manual Count Ln 1
120 100
80 60 40 20
0
Figure 35: Time Series Plot for Manual Counts and Signal Detector Counts at Intersection of SR 237 and SR 141 for Lane 1
160 Signal Count Ln 2
140 Manual Count Ln 2
120 100
80 60 40 20
0
Figure 36: Time Series Plot for Manual Counts and Signal Detector Counts at Intersection of SR 237 and SR 141for Lane 2
36

3/18/2014 7:30 3/18/2014 7:45 3/18/2014 8:00 3/18/2014 8:15 3/18/2014 8:30 3/18/2014 8:45 3/18/2014 9:00 3/18/2014 9:15 3/18/2014 9:30 3/18/2014 9:45 3/18/2014 10:00 3/18/2014 10:15 3/18/2014 10:30

3/18/2014 7:30 3/18/2014 7:45 3/18/2014 8:00 3/18/2014 8:15 3/18/2014 8:30 3/18/2014 8:45 3/18/2014 9:00 3/18/2014 9:15 3/18/2014 9:30 3/18/2014 9:45 3/18/2014 10:00 3/18/2014 10:15 3/18/2014 10:30

160 Signal Count Ln 3
140 Manual Count Ln 3
120 100
80 60 40 20
0

3/18/2014 7:30 3/18/2014 7:45 3/18/2014 8:00 3/18/2014 8:15 3/18/2014 8:30 3/18/2014 8:45 3/18/2014 9:00 3/18/2014 9:15 3/18/2014 9:30 3/18/2014 9:45 3/18/2014 10:00 3/18/2014 10:15 3/18/2014 10:30

Figure 37: Time Series Plot for Manual Counts and Signal Detector Counts at Intersection of SR 237 and SR 141 for Lane 3

500

450

Signal Count

400

Manual Count (3 lanes)

350

Manual Count (All Lanes)

300

250

200

150

100

50

0

3/18/2014 7:30 3/18/2014 7:45 3/18/2014 8:00 3/18/2014 8:15 3/18/2014 8:30 3/18/2014 8:45 3/18/2014 9:00 3/18/2014 9:15 3/18/2014 9:30 3/18/2014 9:45 3/18/2014 10:00 3/18/2014 10:15 3/18/2014 10:30

Figure 38: Time Series Plot for Manual Counts and Signal Detector Counts at Intersection of SR 237 and SR 141 for 3 Through-Lanes and All 5 Lanes Including Left Turn Lanes
37

4.8 Summary of Results
The results of the comparative tests at all the study sites are provided in Table 2. Table 2 provides, at a lane-by-lane level, the mean percentage errors, the 95% confidence interval around the mean percentage errors, the mean absolute percentage errors and root mean square percentage errors. The following are the major observations:
The pneumatic tube counts, which were used as the baseline counts for the rest of the comparisons, were within 10% of the manual counts at lane-by-lane level comparison of 15-minute aggregates. The manual counts were treated as ground truth in this study.
Intersection signal detector counts matched closely with the tube counts, with the 95% confidence intervals of the mean percentage errors within 10%. The errors increased significantly under high volume conditions where is it likely that the vehicle queues extended over the induction loop detection zone, causing over-counting under these conditions.
One of the sites, SR 141 at East Jones Bridge, showed high differences between the intersection signal counts and the pneumatic tube counts. Comparison with the upstream and downstream detectors (time series plots available in the appendix) indicated that the adjacent intersection signal counts matched quite well between themselves but were all different from the pneumatic tube counts. Since the validity of the pneumatic tube counts at this site could not be verified after the fact, conclusions were not drawn based on data at this site. The magnitude of difference has been provided in the summary table, but should not be interpreted as error level for the intersection signal detector data in this instance.
Higher level of variability of differences was observed in the lane level counts than the aggregate over all lanes counts. Part of this can be explained by the possible lane changes between the spatially separated detection zones of the pneumatic tube counters and the intersection signal detectors.
Detector malfunctions were noticed at some locations such as the SR 140 and Fouts Road intersection
It was observed that in several cases, the detection zones of the intersection signal detectors were downstream of the beginning of the turn lanes. It was confirmed in the comparisons that this causes severe undercounting in these scenarios because the turning vehicles are not accounted for.
38

Table 2: Summary of Percentage Errors

Location

Direction Lane

SR 140 @ River Exchange Dr EB

Fast

SR 140 @ River Exchange Dr EB

Slow

SR 140 @ River Exchange Dr EB

All

SR 140 @ River Exchange Dr WB

Fast

SR 140 @ River Exchange Dr WB

Slow

SR 140 @ River Exchange Dr WB

All

SR 140 @ River Exchange Dr EB

Fast

SR 140 @ River Exchange Dr EB

Slow

SR 140 @ River Exchange Dr EB

All

SR 92 @ Woodlands Rd

EB

Fast

SR 92 @ Woodlands Rd

EB

Slow

SR 92 @ Woodlands Rd

EB

All

SR 92 @ Woodlands Rd

WB

Fast

SR 92 @ Woodlands Rd

WB

Slow

SR 92 @ Woodlands Rd

WB

All

SR 140 @ Steeplechase Rd EB

Fast

SR 140 @ Steeplechase Rd EB

Slow

SR 140 @ Steeplechase Rd EB

All

Evaluated Count MPE Lower 95% Upper 95% MAPE RMSPE

Signal

1.9% 4.0%

-0.2%

8.0% 14.8%

Signal

-5.2% -4.0%

-6.4%

5.4% 10.1%

Signal

-4.3% -1.3%

-7.3%

3.6% 9.9%

Signal

4.4% 5.4%

3.4%

5.6% 8.3%

Signal

-9.7% -7.8%

-11.6% 9.8% 16.6%

Signal

-7.6% -5.4%

-9.7%

3.7% 7.6%

Tube

3.1% -2.1%

8.4%

4.9% 7.3%

Tube

2.3% 0.1%

4.4%

2.7% 3.5%

Tube

2.7% 0.7%

4.7%

9.7% 3.7%

Signal

7.4% 8.8%

6.1%

7.5% 10.0%

Signal

-3.6% -2.8%

-4.5%

3.9% 5.5%

Signal

2.0% 2.6%

1.5%

2.2% 3.3%

Signal

-1.5% 0.0%

-3.0%

4.1% 7.5%

Signal

-1.1% 0.2%

-2.3%

4.0% 6.2%

Signal

-1.3% -0.2%

-2.3%

2.3% 5.3%

Signal

-10.0% -6.2%

-13.8%

11.1% 21.3%

Signal

-0.5% 0.7%

-1.7%

4.1% 8.4%

Signal

-5.7% -3.2%

-8.3%

7.6% 13.9%

39

Location

Direction Lane Evaluated Count MPE Lower 95% Upper 95% MAPE RMSPE

SR 140 @ Steeplechase Rd WB

Fast Signal

2.9% 3.9%

1.9%

3.7% 7.7%

SR 140 @ Steeplechase Rd WB

Slow Signal

-2.7% -1.9%

-3.4%

3.1% 6.0%

SR 140 @ Steeplechase Rd WB

All Signal

0.7% 1.4%

0.0%

2.4% 3.7%

SR141 @ E. Jones Bridge Rd* NB

Fast Signal

-23.3% -17.8%

-28.8%

23.5% 36.0%

SR141 @ E. Jones Bridge Rd* NB

Slow Signal

-24.5% -15.4%

-33.7%

24.6% 51.7%

SR141 @ E. Jones Bridge Rd* NB

All Signal

-6.2% -5.1%

-7.3%

6.2% 8.2%

SR141 @ E. Jones Bridge Rd* SB

Fast Signal

24.9% 30.6%

19.1%

30.0% 38.0%

SR141 @ E. Jones Bridge Rd* SB

Slow Signal

-27.2% -20.9%

-33.5%

30.4% 41.4%

SR141 @ E. Jones Bridge Rd* SB

All Signal

-20.1% -15.3%

-24.8%

21.6% 31.0%

SR 314 @ Banks Rd

NB

Fast Signal

-1.4% 0.7%

-3.6%

6.2% 15.4%

SR 314 @ Banks Rd

NB

Slow Signal

-3.4% -2.1%

-4.7%

4.3% 9.7%

SR 314 @ Banks Rd

NB

All Signal

0.5% 0.8%

0.2%

1.4% 2.1%

SR 314 @ Banks Rd

SB

Fast Signal

6.2% 8.1%

4.3%

7.3% 11.8%

SR 314 @ Banks Rd

SB

Slow Signal

-2.0% -1.5%

-2.6%

2.1% 3.5%

SR 314 @ Banks Rd

SB

All Signal

-0.1% 0.4%

-0.6%

1.6% 2.5%

SR 141 @ Old Alabama Rd NB

All Signal

-2.6% -1.6%

-3.7%

2.1% 4.1%

SR 141 @ Old Alabama Rd SB

All Signal

0.7% 1.4%

0.0%

2.1% 4.1%

SR 92@ Roswell Crossing EB

Fast Signal

1.3% 2.3%

0.4%

4.2% 6.6%

SR 92@ Roswell Crossing EB

Middle Signal

-9.6% -8.4%

-10.8%

9.7% 12.9%

SR 92@ Roswell Crossing EB

Slow Signal

24.3% 27.9%

20.6%

24.3% 35.4%

SR 92@ Roswell Crossing EB

All Signal

4.3% 5.3%

3.3%

5.0% 8.3%

40

Location SR 92@ Roswell Crossing SR 92@ Roswell Crossing SR 92@ Roswell Crossing SR 92@ Roswell Crossing SR 92@ Roswell Crossing SR 92@ Roswell Crossing SR 92@ Roswell Crossing SR 92@ Roswell Crossing SR 141 @ SR 237 SR 141 @ SR 237 SR 141 @ SR 237 SR 141 @ SR 237 SR 92 @ Woodstock SR 92 @ Woodstock SR 92 @ Woodstock SR 92 @ Woodstock SR 92 @ Woodstock SR 92 @ Woodstock SR 92 @ Woodstock SR 92 @ Woodstock SR 92@ Bowen Road

Direction Lane Evaluated Count MPE Lower 95% Upper 95% MAPE RMSPE

EB

Fast Signal D/S

-2.1% -0.1%

-4.1%

11.0% 13.9%

EB

Middle Signal D/S

-10.1% -7.5%

-12.6%

15.3% 20.5%

EB

Slow Signal D/S

21.7% 25.2%

18.2%

22.1% 32.6%

EB

All Signal D/S

2.1% 3.2%

1.1%

4.9% 7.4%

WB

Fast Signal

-4.1% 0.0%

-8.2%

9.3% 28.7%

WB

Middle Signal

2.5% 3.2%

1.9%

4.0% 5.2%

WB

Slow Signal

0.3% 0.8%

-0.1%

2.1% 3.3%

WB

All Signal

-0.5% 0.9%

-2.0%

4.0% 10.0%

EB

Slow Signal/VDS

11.8% 0.7%

22.9%

17.9% 19.9%

EB

Middle Signal/VDS

9.6% 0.2%

18.9%

11.1% 18.5%

EB

Fast Signal/VDS

2.6% 1.0%

4.2%

3.1% 3.8%

EB

All Signal/VDS

6.1% 3.9%

8.3%

6.1% 7.1%

WB

Fast Tube

4.5% 2.6%

6.4%

4.5% 5.5%

WB

Middle Tube

-6.0% -8.4%

-3.7%

6.5% 7.1%

WB

Slow Tube

8.7% 4.3%

13.2%

8.9% 11.3%

WB

All Tube

3.0% 1.8%

4.2%

3.0% 3.5%

EB

Fast Signal

12.9% 19.2%

6.6%

19.1% 46.0%

EB

Middle Signal

58.1% 72.7%

43.6%

58.5% 118.0%

EB

Slow Signal

-8.0% -3.9%

-12.1%

20.5% 30.0%

EB

All Signal

19.6% 25.1%

14.0%

19.6% 21.8%

EB

Fast Signal

6.4% 8.9%

3.9%

12.6% 18.7%

41

Location SR 92@ Bowen Road SR 92@ Bowen Road SR 92@ Bowen Road

Direction Lane Evaluated Count MPE Lower 95% Upper 95% MAPE RMSPE

EB

Middle Signal

19.1% 22.0%

16.2%

20.0% 27.9%

EB

Slow Signal

4.4% 6.6%

2.1%

11.1% 16.3%

EB

All Signal

11.9% 14.0%

9.8%

13.4% 18.9%

* Possibly faulty baseline counts

42

5.0 Conclusions and Recommendations
5.1 Findings
The objective of this study was to perform an evaluation of the feasibility of integrating intersection traffic signal data into a traffic monitoring program. About six hundred and eighty hours of directional data across nine study sites around Metro Atlanta was analyzed at a laneby-lane level in this study. The following are the major findings and related observations:
Accuracy of 15-minute aggregates of vehicle counts is above 90% at a 95% level of confidence at majority of the study sites.
Data quality at some intersections (SR 92 at Woodstock Rd) are affected by insufficient offset of the detection zone from the stop bar, leading to queuing of vehicles over the detector and resulting in over-counting of vehicles.
Results of tests at other high volume intersections did not reveal any severe quality issues as observed at the Woodstock Rd intersection. The quality of data varies from site to site and nature of installation.
Data from intersections using video based detection are of comparable quality as other intersections using inductive loop detectors. Like other technologies, video based detection is sensitive to calibration. At the video detection test site two out of three lanes reported data with less than 10% errors (MPE) while one of the lanes reported a mean percentage error of 12%. While it was not verified via a re-calibration and reevaluation in this study, it is likely that the detection accuracy could be improved by better placement of the detection zone in the camera view.
5.2 Intersection Eligibility Criteria
On GDOT's Regional Traffic Operations corridors, the signals had detectors installed only on the major roads. No volumes are available from minor roads. Based on this information alone, it is not possible to accurately determine the volumes downstream of an intersection. Therefore the volumes at a single screenline (midblock location) cannot be generated based on the detectors at a single intersection. It is necessary to use a pair of adjacent intersection's detectors, preferably with no other cross-streets in between, to estimate a midblock screenline count (see Figure 39)

Offset Loop used for WB lane

Portable data collection

Offset Loop used for EB lane detector
Figure 39: Detection Layout Schematic
43

For using the intersection signal detector data for traffic monitoring the following caution need to be exercised:
Discard locations with data spikes i.e. detectors reporting more than 500 vehicles per lane per 15 minutes (threshold can be fine-tuned) at any single data-point (See Figure 40)
Ensure stability of data connection for completeness of data at remote locations which use wireless data transfers
Data availability does not guarantee data quality. Check plot of historical data for expected traffic patterns before considering use of data in traffic monitoring (see Figure 41)
Use data from signal detector downstream of midblock location (i.e. use data from two intersections instead of one) (see Figure 39)
Use data from locations where the inductive loop detectors are physically located upstream of the beginning of turn lanes (if any) (see Figure 42)
Avoid locations where there is evidence (can be confirmed with site visit during peak period) of vehicles queuing over the detection zone

1400 1200 1000
800 600 400 200
0

Data with spikes

12:00 0:00
12:00 0:00
12:00

12:00 0:00
12:00 0:00
12:00

350 300 250 200 150 100
50 0
Figure 40: Example of Traffic Counts with and without Data Spikes
Note the different scales in the figures 44

Data with normal variations

5/4/2014 0:00 5/5/2014 0:00 5/6/2014 0:00 5/7/2014 0:00 5/8/2014 0:00 5/9/2014 0:00 5/10/2014 0:00 5/11/2014 0:00

350
300
250
200
150
100
50
0
Figure 41: Example of Time Series Plot with well Defined Traffic Volume Pattern
5.3 Ideal Intersection for Data Use
Based on the above eligibility criteria the following are the characteristics of a site that would be ideal for use in traffic monitoring: Detector data does not display any sudden spikes on a regular basis and there is no
known occurrence of queuing of vehicles over the detector. Here a spike is defined as the existence of a few data points within a series that deviate significantly from the other data points. At least one week of data should be used to generate time series plots of the data to visually check for data spikes. The signal cabinet communicates with the central server via a stable wired connection. Wireless connections have been found to be unreliable, and may not guarantee completeness of data. Data plots agree with expected increases and decreases in traffic demand during peak and off peak travel periods. At least one typical week of data should be used to generate time series plots of the data to visually check for expected data patterns (see Figure 41). Here a typical week is defined as a week that does not have any holidays, weather events etc. and is not adjacent to long weekends. Data is available from two adjacent intersections. Data from signal detector downstream of midblock location will be used for traffic monitoring (see Figure 39).
45

The intersection signal's inductive loop detectors are physically located upstream of the beginning of turn lanes (if any) (see Figure 42)

Beginnning of turn lane

Existing Detection Zone

Proposed Detection Zone

Figure 42: Proposed Detection Zone Placement Upstream of Beginning of Turn Lanes (Image Courtesy: Google Maps)
5.4 Recommendations
The comparison results in this study indicate that in majority of the cases, the intersection signal detector data is similar in quality to pneumatic tube count data in terms of both the mean and variability of the errors. While there is variability in the level of error from site to site, in general following the intersection eligibility criteria guidelines should help ensure high quality of data for use in the traffic monitoring program. If the data from a particular site is expected to be used extensively on a long term basis, validation of the data via short term counts is recommended.
The intersection signal detectors do not provide vehicle classification information. Therefore the intersection signal detectors cannot be expected to replace the permanent traffic counting stations used in the traffic monitoring program. However, these detectors provide continuous data throughout the year and can possibly be used for other purposes such as determining seasonality factors.
5.4.1 Installation Considerations
To improve the usability of intersection signal detector data for traffic monitoring it is recommended that for future installations and maintenance on existing detectors, detection zones are moved further upstream beyond the maximum queue length of a typical peak hour queue and beyond the beginning of the turn lanes. If it is not possible to move detections zones upstream of turn lanes, detectors should be installed on turn lanes as well.
5.4.2 Communication
A significant amount of effort was spent in the early stages of the project to establish the communication connections required to access the data. The signal detector data accessed via
46

SOAP XML requests to the TACTICS server did not provide guaranteed data quality and was therefore deemed unusable for traffic monitoring purposes. The signal detector data is stored on local memory on the controller in the form of 15 minute counts (by-lane). This data is polled out by a central server on a periodic basis (one to two times daily) and pushed into an archive. The TACTICS interface provides a graphical user interface to query out data for specific intersections. However this process is not favorable for automation. For this project, a procedure was set up to automatically query the database (by GDOT's database administrator) and transfer a text file with the queried data to Georgia Tech's server via Secure Copy Protocol (SCP). While this setup is acceptable for a research project with temporary and transient data needs, a stable setup needs to be designed if the data is to be used for traffic monitoring purposes.
In addition there are other communication hurdles that need to be surmounted before the signal detector data can be reliably used in the traffic monitoring program:
The remote site (SR 314) that used communication over cellular network for the periodic polls of the 15 minute archives failed regularly to transfer the entire dataset. This created availability issues in the dataset and prevented the study from obtaining complete 24 hour counts at the site. In the absence of more reliable communications, this issue needs to be addressed by either increasing the frequency of the polls or increasing the aggregation interval length to reduce amount of data transferred per poll.
There are signals in the RTOP corridors that are not connected to the central TACTICS server. Hence the data from these signal detectors are not available in GDOT's data archive. If data from these corridors are to be used in the traffic monitoring program, the data needs to be retrieved periodically and archived by GDOT in a manner similar to the corridors currently connected to the central server.
5.4.3 Future Research
The results of this study indicate that the quality of data varies from site to site. Deployment characteristics, quality of calibration, communication faults, location of loops with respect to the beginning of turn lanes, presence of queues extending over the detection zone etc. are some of the factors that affect the quality of the data as it pertains to the use of this data in the traffic monitoring program. However several of these factors are either theoretically difficult to quantify or is too resource intensive to measure for a large number of sites. Typically, the quality of data at a site can be determined by performing baseline counts near the existing detectors i.e. by looking at the resulting data stream rather than trying to determine the cause of the anomaly. However, such quality checks need to be performed on a periodic basis to ensure that the quality of data has not degraded over time due to changes in any of the above mentioned factors. One of the purposes of the current study was to investigate the feasibility of eliminating short-term counts where signal detector counts are available. Unless the higher quality data requirement is justified by real time traffic operations, performing validation checks would defeat the purpose of resource conservation because it will require more resources for the validation effort than it saves by eliminating temporary counts. Under these resource constraints, it would be helpful to have a mechanism for estimating, in lieu of measuring, the quality of the data without performing extensive field measurements.
Further research needs to be conducted to develop a methodology for estimating data quality based on readily available information and without requiring additional data collection. It is envisioned that a system for assigning "Quality Ratings" for Signal detector data at each intersection with offset detectors can be developed. The quality rating will be based on
47

comparisons of data from roads with similar volumes and demands. The comparison data can be obtained from nearby permanent count stations, or data from Signals detectors from nearby intersections. At the lane level, comparisons can be done with data from adjacent lanes to detect anomalies. Rational rule based checks such as variation of volumes with time of day, location of detection loops with respect to turn lanes etc. can be used along with the data comparison statistics to determine the quality ratings. The quality ratings would be available at the approach level. The quality ratings will provide a ready reference for choosing intersection approaches with acceptable data quality for use in the traffic monitoring program. In addition, this methodology provides a scalable architecture where new installations can be easily integrated in the list of rated intersections, without requiring additional short-term data collection for every new installation site. Once the base data patterns are established, this framework can be leveraged for monitoring data for real-time operations as well, by developing correlation factors and ranges of expected variation that will help identify deviation of data at individual detectors which could be linked to potential degradation in data quality.
48

6.0 References
Sherry L. Skszek, "State-of-the-Art" Report on Non-Traditional Traffic Counting Methods, FHWA-AZ-01-503, October 2001.
NIT Phase II Evaluation of Non-Intrusive Technologies for Traffic Detection, Final Report, Minnesota Department of Transportation, St. Paul, MN, September 2002.
Field Test of Monitoring of Urban Vehicle Operations Using Non-intrusive Technologies Final Report, Minnesota Department of Transportation, SRF Consulting Group, Inc., MN, May 1997
Middleton D., Jasek D., and Parker R. "Evaluation of Some Existing Technologies for Vehicle Detection," Project Summary Report 1715-S. Texas Transportation Institute, September 1999.
Nathan A. Weber, Verification of Radar Vehicle Detection Equipment, Report SD98-15F, March 1999
McGowen, P. and Sanderson, M. Accuracy of Pneumatic Road Tubes. Institute of Transportation Engineers Western District Annual Meeting. Anchorage, AK. July 2011.
"Using Existing Loops at Signalized Intersections for Traffic Counts: An ITE Informational Report", ITE Journal, Vol: 78, Issue: 2, February 2008
49

7.0 Appendix

List of Figures (Appendix)
Figure A-1: Time Series Plots for Pneumatic Tube Counts and Signal Detector Counts Aggregated over All Lanes at Intersection of SR 141 and Old Alabama Road for NB (Top) and SB (Bottom) (Note: Lane-by-lane data was not provided for this site for the intersection signal data) ............................................................................................................................................ 7-5
Figure A-2: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 1..................................................... 7-6
Figure A-3: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 2..................................................... 7-6
Figure A-4: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 3..................................................... 7-7
Figure A-5: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Aggregates over All Lanes..................... 7-7
Figure A-6: Y-Y Plots for Pneumatic Tube Counts Vs Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Lane 1 (Top-Left), Lane 2 (Top-Right), Lane 3 (BottomLeft) and Aggregates over All Lanes (Bottom Right)................................................................ 7-8
Figure A-7: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Lane 1 ............ 7-9
Figure A-8: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Lane 2 ............ 7-9
Figure A-9: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Lane 3 .......... 7-10
Figure A-10: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Aggregates over All Lanes .................................................................................................................................. 7-10
Figure A-11: Y-Y Plots for Pneumatic Tube Counts Vs Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Lane 1 (Top-Left), Lane 2 (Top-Right), Lane 3 (BottomLeft) and Aggregates over All Lanes (Bottom Right).............................................................. 7-11
Figure A-12: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 EB and River Exchange Parkway for Lane 1 712
Figure A-13: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 EB and River Exchange Parkway for Lane 2 713
Figure A-14: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Lane 1 .................................................................................................................................................. 7-13
Figure A-15: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Lane 2 .................................................................................................................................................. 7-14
7-1

Figure A-16: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 EB and River Exchange Parkway for Aggregates over All Lanes ....................................................................................................... 7-14
Figure A-17: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Aggregates over All Lanes ....................................................................................................... 7-15
Figure A-18: Time Series Plot for Manual Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Aggregates over All Lanes .......................................................................................................................... 7-15
Figure A-19: Time Series Plot for Manual Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Aggregates over All Lanes .......................................................................................................................... 7-16
Figure A-20: Time Series Plot for Manual Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Aggregates over All Lanes .......................................................................................................................... 7-16
Figure A-21: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 1.............................................................. 7-17
Figure A-22: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 2.............................................................. 7-17
Figure A-23: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 3.............................................................. 7-18
Figure A-24: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Aggregates over All Lanes.............................. 7-18
Figure A-25: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Aggregates over All Lanes with Count Ceiling per lane at 350 vehicles .................................................................................................................. 7-19
Figure A-26: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Lane 1 ............................................................... 7-19
Figure A-27: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Lane 2 ............................................................... 7-20
Figure A-28: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Lane 3 ............................................................... 7-20
Figure A-29: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Aggregates over All Lanes ............................... 7-21
Figure A-30: Time Series Plot for Manual Counts, Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 1 ............................................. 7-21
Figure A-31: Time Series Plot for Manual Counts, Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 2 ............................................. 7-22
Figure A-32: Time Series Plot for Manual Counts, Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 3 ............................................. 7-22
7-2

Figure A-33: Time Series Plot for Manual Counts, Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Aggregates over All Lanes ............. 7-23
Figure A-34: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 EB and Steeple ChaseDr for Lane 1.................................................... 7-24
Figure A-35: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 EB and Steeple ChaseDr for Lane 2.................................................... 7-24
Figure A-36: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 WB and Woodlands Parkway for Lane 1 ............................................ 7-25
Figure A-37: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 WB and Woodlands Parkway for Lane 2 ............................................ 7-25
Figure A-38: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 EB and Steeple Chase Drive for Aggregates over All Lanes.............. 7-26
Figure A-39: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 WB and Woodland Parkway for Aggregates over All Lanes ............. 7-26
Figure A-40: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 NB and Banks Rd for Lane 1 .............................................................. 7-27
Figure A-41: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 NB and Banks Rd for Lane 2 .............................................................. 7-27
Figure A-42: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 SB and Creek Wood Trl for Lane 1 .................................................... 7-28
Figure A-43: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 SB and Creek Wood Trl for Lane 2 .................................................... 7-28
Figure A-44: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 EB and Banks Rd for Aggregates over All Lanes............................... 7-29
Figure A-45: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 WB and Creek Wood Trl for Aggregates over All Lanes ................... 7-29
Figure A-46: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 EB and SR 237 for Lane 1................................................................... 7-30
Figure A-47: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 EB and SR 237 for Lane 2................................................................... 7-30
Figure A-48: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 WB and SR 237 for Lane 3 ................................................................. 7-31
Figure A-49: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 WB and SR 237 for Aggregates over All Lanes ................................. 7-31
Figure A-50: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 NB and East Jones Bridge Road for Lane 1 ........................................ 7-32
Figure A-51: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 NB and East Jones Bridge Road for Lane 2 ........................................ 7-32
Figure A-52: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 SB and Forum Drive for Lane 1 .......................................................... 7-33
7-3

Figure A-53: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 SB and Forum Drive for Lane 2 .......................................................... 7-33 Figure A-54: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 NB and East Jones Bridge Road for Aggregates over All Lanes ........ 7-34 Figure A-55: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 SB and Forum Drive for Aggregates over All Lanes .......................... 7-34 Figure A-56: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Dr for Lane 1................................................................... 7-35 Figure A-57: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Dr for Lane 2................................................................... 7-35 Figure A-58: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Woodlands Parkway for Lane 1 .............................................. 7-36 Figure A-59: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Woodlands Parkway for Lane 2 .............................................. 7-36 Figure A-60: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Drive for Aggregates over All Lanes .............................. 7-37 Figure A-61: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Woodland Parkway for Aggregates over All Lanes ............... 7-37 Figure A-62: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Bowen Road for Lane 1 .......................................................... 7-38 Figure A-63: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Bowen Road for Lane 2 .......................................................... 7-38 Figure A-64: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Bowen Road for Lane 3 .......................................................... 7-39 Figure A-65: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Bowen Road for Aggregates over All Lanes .......................... 7-39
7-4

7.1 SR 141 at Old Alabama Road

800 Tube Count FL+ML+SL

700

Tube Count FL + ML

NB Signal Count 600

500

400

300

200

100

0

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 2/18/2014 10:00 2/18/2014 12:30 2/18/2014 15:00 2/18/2014 17:30 2/18/2014 20:00 2/18/2014 22:30

800

Tube Count FL + ML + SL

700

Tube Count FL +ML

600

Signal Count SB

500

400

300

200

100

0

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 2/18/2014 10:00 2/18/2014 12:30 2/18/2014 15:00 2/18/2014 17:30 2/18/2014 20:00 2/18/2014 22:30

Figure A-1: Time Series Plots for Pneumatic Tube Counts and Signal Detector Counts Aggregated over All Lanes at Intersection of SR 141 and Old Alabama Road for NB (Top) and SB (Bottom) (Note: Lane-by-lane data was not provided for this site for the intersection signal data)
7-5

7.2 SR 92 at Roswell Crossing
350 300 250 200 150 100
50 0

Tube Count 92 @ RosCr EB1

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-2: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 1

350 Tube Count

300

92 @ RosCr EB2

250

200

150

100

50

0

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-3: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 2
7-6

350 Tube Count

300

92 @ RosCr EB3

250

200

150

100

50

0

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-4: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Lane 3

1000 900 800 700 600 500 400 300 200 100 0

Tube Count 92 @ RosCr EB All Ln

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-5: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Roswell Crossing for Aggregates over All Lanes
7-7

500

500

450

450

400

400

350

350

300

300

250

250

200

200

150

150

100

100

50

50

0

0

500

450

400

300 200
400 250
500 300
600 350
700

200 150

100 100

50

0

500

450

400 450

300 300
350 350
400

250 250

200

150

100

50

0

500

1000

450

900

400

800

350

700

300

600

250

500

200

400

150

300

100

200

50

100

0

0

1000

900

800

0

500

200

150

100

50

0

Figure A-6: Y-Y Plots for Pneumatic Tube Counts Vs Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Lane 1 (Top-Left), Lane 2 (Top-Right), Lane 3 (Bottom-Left) and Aggregates over All Lanes (Bottom Right)

7-8

350 Tube Count D/S

300

92 @ RosCr EB1

250

200

150

100

50

0

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-7: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Lane 1

350 Tube Count D/S

300

92 @ RosCr EB2

250

200

150

100

50

0

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-8: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Lane 2
7-9

350

Tube Count D/S

300

92 @ RosCr EB3

250

200

150

100

50

0

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-9: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Lane 3

1000 900 800 700 600 500 400 300 200 100 0

Tube Count D/S 92 @ RosCr EB All Ln

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-10: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Aggregates over All Lanes
7-10

500 450 400 350 300 250 200 150 100
50 0 0 50 100 150 200 250 300 350 400 450 500
500 450 400 350 300 250 200 150 100
50 0 0 50 100 150 200 250 300 350 400 450 500

500 450 400 350 300 250 200 150 100
50 0 0 50 100 150 200 250 300 350 400 450 500
1000 900 800 700 600 500 400 300 200 100 0 0 1002003004005006007008009001000

Figure A-11: Y-Y Plots for Pneumatic Tube Counts Vs Signal Detector Counts at Intersection of SR 92 EB and Roswell Crossing for Lane 1 (Top-Left), Lane 2 (Top-Right), Lane 3 (Bottom-Left) and Aggregates over All Lanes (Bottom Right)

7-11

5/5/2014 12:00 5/5/2014 16:48 5/5/2014 21:36
5/6/2014 2:24 5/6/2014 7:12 5/6/2014 12:00 5/6/2014 16:48 5/6/2014 21:36 5/7/2014 2:24 5/7/2014 7:12 5/7/2014 12:00

7.3 SR 140 at River Exchange Parkway
500 450 400 350 300 250 200 150 100
50 0

Tube Count Signal Count EB Ln 1

Figure A-12: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 EB and River Exchange Parkway for Lane 1

500 Tube Count

450

Signal Count EB Ln 2

400

350

300

250

200

150

100

50

0

5/5/2014 12:00 5/5/2014 16:48 5/5/2014 21:36
5/6/2014 2:24 5/6/2014 7:12 5/6/2014 12:00 5/6/2014 16:48 5/6/2014 21:36 5/7/2014 2:24 5/7/2014 7:12 5/7/2014 12:00

7-12

Figure A-13: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 EB and River Exchange Parkway for Lane 2

500 Tube Count

450

Signal Count WB Ln 1

400

350

300

250

200

150

100

50

0

Figure A-14: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Lane 1

5/5/2014 12:00 5/5/2014 16:48 5/5/2014 21:36
5/6/2014 2:24 5/6/2014 7:12 5/6/2014 12:00 5/6/2014 16:48 5/6/2014 21:36 5/7/2014 2:24 5/7/2014 7:12 5/7/2014 12:00

7-13

500 Tube Count

450

Signal Count WB Ln 2

400

350

300

250

200

150

100

50

0

5/5/2014 12:00 5/5/2014 16:48 5/5/2014 21:36
5/6/2014 2:24 5/6/2014 7:12 5/6/2014 12:00 5/6/2014 16:48 5/6/2014 21:36 5/7/2014 2:24 5/7/2014 7:12 5/7/2014 12:00

Figure A-15: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Lane 2

900 Tube Count EB All Ln

800

Signal Count EB All Ln

700

600

500

400

300

200

100

0

5/5/2014 12:00 5/5/2014 16:48 5/5/2014 21:36
5/6/2014 2:24 5/6/2014 7:12 5/6/2014 12:00 5/6/2014 16:48 5/6/2014 21:36 5/7/2014 2:24 5/7/2014 7:12 5/7/2014 12:00

Figure A-16: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 EB and River Exchange Parkway for Aggregates over All Lanes
7-14

900 Tube Count WB All Ln

800

Signal Count WB All Ln

700

600

500

400

300

200

100

0

5/5/2014 12:00 5/5/2014 16:48 5/5/2014 21:36
5/6/2014 2:24 5/6/2014 7:12 5/6/2014 12:00 5/6/2014 16:48 5/6/2014 21:36 5/7/2014 2:24 5/7/2014 7:12 5/7/2014 12:00

Figure A-17: Time Series Plot for Pneumatic Tube Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Aggregates over All Lanes

450 Manual Ln 1
400 Tube Ln 1
350

300

250

200

150

100

50

0

7:12

7:26

7:40

7:55

8:09

8:24

8:38

8:52

9:07

Figure A-18: Time Series Plot for Manual Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Aggregates over All Lanes

7-15

450

Manual Ln 2

400

Tube Ln 2

350

300

250

200

150

100

50

0

7:12

7:26

7:40

7:55

8:09

8:24

8:38

8:52

9:07

Figure A-19: Time Series Plot for Manual Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Aggregates over All Lanes

900

800

700

600

500
Manual All Ln 400
Tube All Ln 300

200

100

0

7:12

7:26

7:40

7:55

8:09

8:24

8:38

8:52

9:07

Figure A-20: Time Series Plot for Manual Counts Downstream of Intersection and Signal Detector Counts at Intersection of SR 140 WB and River Exchange Parkway for Aggregates over All Lanes

7-16

7.4 SR 92 at King Road / Woodstock Road
1200 1000
800 600 400 200
0

Tube + LTL 92 @ King WB1

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure A-21: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 1

1200 1000
800 600 400 200
0

Tube Count 92 @ King WB2

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure A-22: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 2
7-17

1200 1000
800 600 400 200
0

Tube + RTL 92 @ King WB3

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure A-23: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 3

3000 2500 2000 1500 1000
500 0

Tube Count All Ln 92 @ King WB All Ln

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure A-24: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Aggregates over All Lanes
7-18

3000 2500 2000 1500 1000
500 0

Tube Count All Ln
92 @ King WB All Ln Ceiling 350

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure A-25: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Aggregates over All Lanes with Count Ceiling per lane at 350 vehicles

1200 1000
800 600 400 200
0

Tube Count SR 92 EB Ln 1 @ King

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure A-26: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Lane 1
7-19

1200 1000
800 600 400 200
0

Tube Count SR 92 EB Ln 2 @ King

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure A-27: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Lane 2

1200 1000
800 600 400 200
0

Tube Count SR 92 EB Ln 3 @ King

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure A-28: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Lane 3
7-20

1800 1600 1400 1200 1000
800 600 400 200
0

Tube Count SR 92 EB All Ln @ King

10/29/2013 12:00 10/29/2013 18:00
10/30/2013 0:00 10/30/2013 6:00 10/30/2013 12:00 10/30/2013 18:00 10/31/2013 0:00 10/31/2013 6:00 10/31/2013 12:00

Figure A-29: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and King Road for Aggregates over All Lanes

1000 900 800 700 600 500 400 300 200 100 0

Tube + LTL
92 @ King WB1 (Cabinet) Manual WB1+LTL

10/29/2013 16:30 10/29/2013 16:45 10/29/2013 17:00 10/29/2013 17:15 10/29/2013 17:30 10/29/2013 17:45 10/29/2013 18:00 10/29/2013 18:15 10/29/2013 18:30 10/29/2013 18:45 10/29/2013 19:00 10/29/2013 19:15 10/29/2013 19:30

Figure A-30: Time Series Plot for Manual Counts, Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 1
7-21

1000 900 800 700 600 500 400 300 200 100 0

Tube Count
92 @ King WB2 (Cabinet) Manual WB2

10/29/2013 16:30 10/29/2013 16:45 10/29/2013 17:00 10/29/2013 17:15 10/29/2013 17:30 10/29/2013 17:45 10/29/2013 18:00 10/29/2013 18:15 10/29/2013 18:30 10/29/2013 18:45 10/29/2013 19:00 10/29/2013 19:15 10/29/2013 19:30

Figure A-31: Time Series Plot for Manual Counts, Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 2

1000 900 800 700 600 500 400 300 200 100 0

Tube + RTL
92 @ King WB3 (Cabinet) Manual WB3+RTL

10/29/2013 16:30 10/29/2013 16:45 10/29/2013 17:00 10/29/2013 17:15 10/29/2013 17:30 10/29/2013 17:45 10/29/2013 18:00 10/29/2013 18:15 10/29/2013 18:30 10/29/2013 18:45 10/29/2013 19:00 10/29/2013 19:15 10/29/2013 19:30

Figure A-32: Time Series Plot for Manual Counts, Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Lane 3
7-22

2500 2000 1500 1000
500 0

Tube WB All Ln Signal WB All Ln Manual WB All Ln

10/29/2013 16:30 10/29/2013 16:45 10/29/2013 17:00 10/29/2013 17:15 10/29/2013 17:30 10/29/2013 17:45 10/29/2013 18:00 10/29/2013 18:15 10/29/2013 18:30 10/29/2013 18:45 10/29/2013 19:00 10/29/2013 19:15 10/29/2013 19:30

Figure A-33: Time Series Plot for Manual Counts, Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and King Road for Aggregates over All Lanes

7-23

7.5 SR140 at Steeple Chase Drive
400 350 300 250 200 150 100
50 0

Tube Count Signal Count SR 92 EB Ln 1 @ Steeplechase US Signal (Fouts) DS Signal (Champions Green)

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 2/18/2014 10:00 2/18/2014 12:30 2/18/2014 15:00 2/18/2014 17:30 2/18/2014 20:00 2/18/2014 22:30

Figure A-34: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 EB and Steeple ChaseDr for Lane 1

400 Tube Count

350

Signal Count SR 92 EB Ln 2 @ Steeplechase

300

US Signal (Fouts)

250

DS Signal (Champions Green)

200

150

100

50

0

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 2/18/2014 10:00 2/18/2014 12:30 2/18/2014 15:00 2/18/2014 17:30 2/18/2014 20:00 2/18/2014 22:30

Figure A-35: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 EB and Steeple ChaseDr for Lane 2
7-24

400 Tube Count

350

Signal Count SR 92 WB Ln 1 @ Steeplechase

300

DS Signal (Fouts)

250

US Signal (Champions Green)

200

150

100

50

0

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 2/18/2014 10:00 2/18/2014 12:30 2/18/2014 15:00 2/18/2014 17:30 2/18/2014 20:00 2/18/2014 22:30

Figure A-36: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 WB and Woodlands Parkway for Lane 1

400 Tube Count

350

Signal Count SR 92 WB Ln 2 @ Steeplechase

300

DS Signal (Fouts)

250

US Signal (Champions Green)

200

150

100

50

0

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 2/18/2014 10:00 2/18/2014 12:30 2/18/2014 15:00 2/18/2014 17:30 2/18/2014 20:00 2/18/2014 22:30

Figure A-37: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 WB and Woodlands Parkway for Lane 2
7-25

600

Tube Count

Signal Count SR 92 EB All Ln @ Steeplechase

500

US Signal (Fouts)

DS Signal (Champions Green) 400

300

200

100

0

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 2/18/2014 10:00 2/18/2014 12:30 2/18/2014 15:00 2/18/2014 17:30 2/18/2014 20:00 2/18/2014 22:30

Figure A-38: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 EB and Steeple Chase Drive for Aggregates over All Lanes

600 Tube Count

Signal Count SR 92 WB All Ln @ Steeplechase

500

DS Signal (Fouts)

US Signal (Champions Green) 400

300

200

100

0

2/18/2014 0:00 2/18/2014 2:30 2/18/2014 5:00 2/18/2014 7:30 2/18/2014 10:00 2/18/2014 12:30 2/18/2014 15:00 2/18/2014 17:30 2/18/2014 20:00 2/18/2014 22:30

Figure A-39: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 140 WB and Woodland Parkway for Aggregates over All Lanes
7-26

7.6 SR 314 at Banks Road and Creek Wood Trail

200

Tube Count

180 Signal Count SR314 NB Ln 1 @ Banks Rd

160 Signal Count SR314 NB Ln 1 @ Creek Wood (U/S)

140

120

100

80

60

40

20

0

5/13/2014 0:00 5/13/2014 4:48 5/13/2014 9:36 5/13/2014 14:24 5/13/2014 19:12 5/14/2014 0:00 5/14/2014 4:48 5/14/2014 9:36 5/14/2014 14:24 5/14/2014 19:12 5/15/2014 0:00

Figure A-40: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 NB and Banks Rd for Lane 1

200

180

Tube Count

160

Signal Count SR314 NB Ln 2 @ Banks Rd

140 Signal Count SR314 NB Ln 2 @ Creek Wood (U/S)
120

100

80

60

40

20

0

Figure A-41: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 NB and Banks Rd for Lane 2
7-27

5/13/2014 0:00 5/13/2014 4:48 5/13/2014 9:36 5/13/2014 14:24 5/13/2014 19:12 5/14/2014 0:00 5/14/2014 4:48 5/14/2014 9:36 5/14/2014 14:24 5/14/2014 19:12 5/15/2014 0:00

200

180

Tube Count

160

Signal Count SR314 SB Ln 1 @ Creek Wood

140 Signal Count SR314 SB Ln 1 @ Banks (U/S)
120

100

80

60

40

20

0

5/13/2014 0:00 5/13/2014 4:48 5/13/2014 9:36 5/13/2014 14:24 5/13/2014 19:12 5/14/2014 0:00 5/14/2014 4:48 5/14/2014 9:36 5/14/2014 14:24 5/14/2014 19:12 5/15/2014 0:00

Figure A-42: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 SB and Creek Wood Trl for Lane 1
200 Tube Count
180 Signal Count SR314 SB Ln 2 @ Creek Wood
160 140 120 100
80 60 40 20
0

5/13/2014 0:00 5/13/2014 4:48 5/13/2014 9:36 5/13/2014 14:24 5/13/2014 19:12 5/14/2014 0:00 5/14/2014 4:48 5/14/2014 9:36 5/14/2014 14:24 5/14/2014 19:12 5/15/2014 0:00

Figure A-43: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 SB and Creek Wood Trl for Lane 2
7-28

5/13/2014 0:00 5/13/2014 4:48 5/13/2014 9:36 5/13/2014 14:24 5/13/2014 19:12 5/14/2014 0:00 5/14/2014 4:48 5/14/2014 9:36 5/14/2014 14:24 5/14/2014 19:12 5/15/2014 0:00

250

Tube Count

200

Signal Count SR314 NB All Ln @ Banks Rd

150

100

50

0

Figure A-44: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 EB and Banks Rd for Aggregates over All Lanes

250 Tube Count

200

Signal Count SR314 SB All Ln @ Creek Wood

150

100

50

0

5/13/2014 0:00 5/13/2014 4:48 5/13/2014 9:36 5/13/2014 14:24 5/13/2014 19:12 5/14/2014 0:00 5/14/2014 4:48 5/14/2014 9:36 5/14/2014 14:24 5/14/2014 19:12 5/15/2014 0:00

Figure A-45: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 314 WB and Creek Wood Trl for Aggregates over All Lanes
7-29

7.7 SR 141 (Peachtree Road) EB at SR 237 (Piedmont Road)

160 Signal Count Ln 1
140

120

Manual Count Ln 1

100

80

60

40

20

0

3/18/2014 7:30 3/18/2014 7:45 3/18/2014 8:00 3/18/2014 8:15 3/18/2014 8:30 3/18/2014 8:45 3/18/2014 9:00 3/18/2014 9:15 3/18/2014 9:30 3/18/2014 9:45 3/18/2014 10:00 3/18/2014 10:15 3/18/2014 10:30

Figure A-46: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 EB and SR 237 for Lane 1

160 Signal Count Ln 2
140

120

Manual Count Ln 2

100

80

60

40

20

0

3/18/2014 7:30 3/18/2014 7:45 3/18/2014 8:00 3/18/2014 8:15 3/18/2014 8:30 3/18/2014 8:45 3/18/2014 9:00 3/18/2014 9:15 3/18/2014 9:30 3/18/2014 9:45 3/18/2014 10:00 3/18/2014 10:15 3/18/2014 10:30

Figure A-47: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 EB and SR 237 for Lane 2
7-30

160 Signal Count Ln 3
140

120

Manual Count Ln 3

100

80

60

40

20

0

3/18/2014 7:30 3/18/2014 7:45 3/18/2014 8:00 3/18/2014 8:15 3/18/2014 8:30 3/18/2014 8:45 3/18/2014 9:00 3/18/2014 9:15 3/18/2014 9:30 3/18/2014 9:45 3/18/2014 10:00 3/18/2014 10:15 3/18/2014 10:30

Figure A-48: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 WB and SR 237 for Lane 3

500

450

Signal Count

400

Manual Count (3 lanes)

350

Manual Count (All Lanes)

300

250

200

150

100

50

0

3/18/2014 7:30 3/18/2014 7:45 3/18/2014 8:00 3/18/2014 8:15 3/18/2014 8:30 3/18/2014 8:45 3/18/2014 9:00 3/18/2014 9:15 3/18/2014 9:30 3/18/2014 9:45 3/18/2014 10:00 3/18/2014 10:15 3/18/2014 10:30

Figure A-49: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 WB and SR 237 for Aggregates over All Lanes
7-31

7.8 SR 141 at East Jones Bridge Rd

400

Tube Count

350

SR 141 NB Ln 1 @ Forum Dr (U/S)

SR 141 NB Ln 1 @ E Jones Bridge Rd

300

SR 141 NB @ Wellington Lake (D/S)

250

200

150

100

50

0

3/4/2014 9:36 3/4/2014 14:24 3/4/2014 19:12
3/5/2014 0:00 3/5/2014 4:48 3/5/2014 9:36 3/5/2014 14:24

Figure A-50: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 NB and East Jones Bridge Road for Lane 1

400

Tube Count

SR 141NB Ln 2 @ Forum Dr (U/S) 350

SR 141 NB Ln 2 @ E Jones Bridge Rd

300

SR 141 NB @ Wellington Lake (D/S)

250

200

150

100

50

0

3/4/2014 9:36 3/4/2014 14:24 3/4/2014 19:12
3/5/2014 0:00 3/5/2014 4:48 3/5/2014 9:36 3/5/2014 14:24

Figure A-51: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 NB and East Jones Bridge Road for Lane 2
7-32

400

Tube Count

350

SR 141 SB Ln 1 @ Forum Dr

SR 141 SB @ E Jones Bridge Rd (U/S) 300
SR 141 SB @ Wellington Lake (2 U/S)

250

200

150

100

50

0

3/4/2014 9:36 3/4/2014 14:24 3/4/2014 19:12
3/5/2014 0:00 3/5/2014 4:48 3/5/2014 9:36 3/5/2014 14:24

Figure A-52: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 SB and Forum Drive for Lane 1

400

Tube Count

350

SR 141 SB Ln 2 @ Forum Dr

SR 141 SB @ E Jones Bridge Rd (U/S)

300

SR 141 SB @ Wellington Lake (2 U/S)

250

200

150

100

50

0

s
Figure A-53: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 SB and Forum Drive for Lane 2
7-33

3/4/2014 9:36 3/4/2014 14:24 3/4/2014 19:12
3/5/2014 0:00 3/5/2014 4:48 3/5/2014 9:36 3/5/2014 14:24

800

Tube Count

SR 141 NB All Ln @ E Jones Bridge Rd

700

SR 141 NB All Ln @ Forum Dr (U/S)

600

SR 141 NB @ Wellington Lake (D/S)

500

400

300

200

100

0

3/4/2014 9:36 3/4/2014 14:24 3/4/2014 19:12
3/5/2014 0:00 3/5/2014 4:48 3/5/2014 9:36 3/5/2014 14:24

Figure A-54: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 NB and East Jones Bridge Road for Aggregates over All Lanes
800 Tube Count
700 SR 141 SB All Ln @ Forum Dr
600 500 400 300 200 100
0

3/4/2014 9:36 3/4/2014 14:24 3/4/2014 19:12
3/5/2014 0:00 3/5/2014 4:48 3/5/2014 9:36 3/5/2014 14:24

Figure A-55: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 141 SB and Forum Drive for Aggregates over All Lanes
7-34

7.9 SR 92 at Dials Drive / Woodlands Parkway
400 350 300 250 200 150 100
50 0

Tube Count SR 92 EB Ln 1 (Dials) DS Signal (Wigley) US Signal (Ragsdale)

2/20/2014 0:00 2/20/2014 2:30 2/20/2014 5:00 2/20/2014 7:30 2/20/2014 10:00 2/20/2014 12:30 2/20/2014 15:00 2/20/2014 17:30 2/20/2014 20:00 2/20/2014 22:30

Figure A-56: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Dr for Lane 1

400

Tube Count

350

SR 92 EB Ln 2 (Dials)

DS Signal (Wigley) 300
US Signal (Ragsdale)

250

200

150

100

50

0

2/20/2014 0:00 2/20/2014 2:30 2/20/2014 5:00 2/20/2014 7:30 2/20/2014 10:00 2/20/2014 12:30 2/20/2014 15:00 2/20/2014 17:30 2/20/2014 20:00 2/20/2014 22:30

Figure A-57: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Dr for Lane 2
7-35

400

Tube Count

350

SR 92 WB Ln 1

(Woodlands)

300

US Signal (Wigley)

250

200

150

100

50

0

2/20/2014 0:00 2/20/2014 2:30 2/20/2014 5:00 2/20/2014 7:30 2/20/2014 10:00 2/20/2014 12:30 2/20/2014 15:00 2/20/2014 17:30 2/20/2014 20:00 2/20/2014 22:30

Figure A-58: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Woodlands Parkway for Lane 1

400

Tube Count

350

SR 92 WB Ln 2

(Woodlands)

300

US Signal (Wigley)

250

200

150

100

50

0

2/20/2014 0:00 2/20/2014 2:30 2/20/2014 5:00 2/20/2014 7:30 2/20/2014 10:00 2/20/2014 12:30 2/20/2014 15:00 2/20/2014 17:30 2/20/2014 20:00 2/20/2014 22:30

Figure A-59: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Woodlands Parkway for Lane 2
7-36

800

Tube Count

700

Signal Count SR 92 EB at Dials

DS Signal (Wigley)

600

US Signal (Ragsdale)

500

400

300

200

100

0

2/20/2014 0:00 2/20/2014 2:30 2/20/2014 5:00 2/20/2014 7:30 2/20/2014 10:00 2/20/2014 12:30 2/20/2014 15:00 2/20/2014 17:30 2/20/2014 20:00 2/20/2014 22:30

Figure A-60: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 EB and Dials Drive for Aggregates over All Lanes

800

Tube Count

700 Signal Count SR

600

92 WB at

Woodlands

500

400

300

200

100

0

2/20/2014 0:00 2/20/2014 2:30 2/20/2014 5:00 2/20/2014 7:30 2/20/2014 10:00 2/20/2014 12:30 2/20/2014 15:00 2/20/2014 17:30 2/20/2014 20:00 2/20/2014 22:30

Figure A-61: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Woodland Parkway for Aggregates over All Lanes
7-37

7.10 SR 92 at Bowen Road
400 350 300 250 200 150 100
50 0

Tube Count SR 92 WB Ln 1 @ Bowen

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-62: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Bowen Road for Lane 1

400

350

300

Tube Count

250

SR 92 WB Ln 2 @ Bowen

200

150

100

50

0

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-63: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Bowen Road for Lane 2
7-38

400

350

Tube + RTL

300

SR 92 WB Ln 3 @ Bowen

250

200

150

100

50

0

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-64: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Bowen Road for Lane 3

1000 900 800 700 600 500 400 300 200 100 0

Tube Count SR 92 WB All Ln @ Bowen

10/29/2013 12:00 10/30/2013 0:00
10/30/2013 12:00 10/31/2013 0:00
10/31/2013 12:00

Figure A-65: Time Series Plot for Pneumatic Tube Counts and Signal Detector Counts at Intersection of SR 92 WB and Bowen Road for Aggregates over All Lanes
7-39

GEORGIA DOT RESEARCH PROJECT 13-10 FINAL REPORT
INTEGRATING INTERSECTION TRAFFIC SIGNAL DATA INTO A TRAFFIC MONITORING PROGRAM
OFFICE OF RESEARCH 15 KENNEDY DRIVE
FOREST PARK, GA 30297-2534