Image reading WEBINAR Lunch Time Tech Automating the Traffic Signal Performance Measures for NJDOT Adaptive Traffic Signal Control Systems

Lunchtime Tech Talk! WEBINAR: Automating Traffic Signal Performance Measures for NJDOT Adaptive Traffic Signal Control Systems

Slide Cover Reading Lunchtime Tech Talk! Automating the Traffic Signal Performance Measures for NJDOT Adaptive Traffic Signal Control Systems - Real-Time Signal Performance Measurement (RT-SPM)

Click For Tech Talk Presentation

The New Jersey Department of Transportation Bureau of Research convened a Lunchtime Tech Talk! Webinar on Automating the Traffic Signal Performance Measures for NJDOT Adaptive Traffic Signal Control Systems on June 29, 2021. The presentation was led by Dr. Peter Jin, of Rutgers-CAIT, Dr. Thomas Brennan, from the College of New Jersey, and Kelly McVeigh from NJDOT’s Mobility Engineering Unit. The three touched upon Phase I research on Real-Time Traffic Signal Performance Measurement and continuing research underway in Phase II  to adapt NJDOT’s existing signaling technology to take advantage of innovative methods in optimizing traffic controls.

Kelly McVeigh, of NJDOT, began the event by introducing the Automated Traffic Signal Performance Measures (ATSPM) and Adaptive Traffic Signal Control Systems (ATSC) concepts. According to McVeigh, Automated Traffic Signal Performance Measures are a suite of measures that help transit agencies to make use of data in optimizing signal timings. ATSPM consists of a dataset of time-stamped events—visually represented through charts—that demonstrate the signal’s performance. For example, how much time the signal is set to green when vehicles are present. The technology, McVeigh said, was “a powerful tool in the toolbox for traffic engineers to monitor performance and even make changes, if agency procedures allow.” ATSPM was first introduced by FHWA as part of the fourth round of the Every Day Counts Initiative (EDC-4).

Slide Reads Challenges with Standard ATSPM Deployment, Standard ATSPm Deployment: High-resolution controllers, data probe and FTP configuration at Signal Boxes. Challenges: Upgrading to high-resolution controllers requires significant investment, $4,000 to $5,000 dollars per intersection. Opportunities: Centralized event logs of Adaptive Signal Control Technology systems. Rapid expansion of ASCT systems. Objectives: Integrate ATSPMs and Adaptive Signal Control Technology )ASCT) systems to produce ATSPM performance metrics. Policies: Dynamically adjust the signal timing in real time in practice. Timing changes (long-term) versus ASCT (real-time/short-term).

In order to avoid costly infrastructure costs of replacing ASCT systems for ATSPM equipment, the researchers devised a method to make use of existing, deployed intersection systems.

McVeigh explained that, while there is already a well-documented system in place to support ASTPM implementation, NJDOT is focusing on adapting existing systems that are already equipped to capture data. Adaptive Traffic Control Systems (ATSC) are installed in nearly 20 percent of NJDOT’s roughly 2,500 signals statewide, and collect data on both traffic controllers and detectors, such as signal performance and vehicle queuing. However, as Dr. Jin then detailed, ATSC data is presently incompatible with ATSPM. In addition, some signals are connected to the centralized network, while others remain isolated. The solution was to develop a means of converting the data, rather than installing new infrastructure.

A team of students from Rutgers, The College of New Jersey, and Rowan University worked with Dr. Jin to bridge data from ATSC to ASTPM. The proposed solution is a program that automatically retrieves traffic controller event logs and then translates them into ASTPM event code, a method that is agnostic to controller type. This allows for a wide variety of data to be collected, and then viewed and optimized using standard ASTPM methods.

Slide image of proposed farmework with a new add on of existign ASCT Sytems going ot get event logs, to ASCT event translator to push ATSPM events, to Database server. The two bullet points read The Newly developed program can automatically retrieve the controller's logfiles and translate records ito standard ATSPM event code. This method is agnostic to the controller type.

The proposed framework would add direct conversion of ASCT events to ATSPM.

Data translation works by taking ATSC logs, such as “Phase Begin Green” stamped with a timecode, and converting that to a numeric code, in this case, “1.” A computer program reads through the SCATS log and assigns certain datapoints to traffic events, such as a gap, which would be coded as “4.” At the conclusion of Phase 1, the team has been able to convert all major events to ASTPM metrics. Going forward, they are working on using geolocated video data to reconstruct stopping data, allowing for more refined information that enables real-time traffic signal adjustments.

Many locations on NJDOT’s network are not properly equipped to convey upstream information on vehicles, particularly during the red phase. The ingenious solution is to locate a “Stop-Bar” within the signal detector that registers when vehicles have begun queuing. This data is then correlated with spatial Google Maps data that precisely locates the vehicles’ position. Information from the Autoscope video-based tracking technology is then used to calculate the vehicle’s trajectory, using the Shockwave Theory of traffic flow. The benefit of such a method is better data on how vehicles approach a red signal, which can then be optimized through ASTPM.

Reads event Translator method, converted signal events will be imported into this ATSPM database. Following this, the ATSPM software can generate performance metrics and produce visualization to suppport maintenance and operations. Table shows signal timing and phase-related event and code used by ATSPMs.

ASTC events are translated using a computer program that can recognize various events and code them as such.

Dr. Brennan then demonstrated the technology in action, sharing his screen to show the ATSPM Server and its variety of tools. He selected a sample intersection, US-1 and Harrison Street near the Millstone River, and brought up a chart showing the Purdue Phase Diagram (PCD). The PCD is a means of graphically representing the number of vehicles passing through an intersection with respect to phase time. In an ideal situation, vehicles should arrive on green, instead of red, when they will have to wait. Another chart represented the traffic split by time of day and duration of the phase. When the technology is fully implemented, such data should be uploaded every 15 minutes, allowing for near real-time monitoring.

From the ATSPM data, Dr. Brennan showed that one signal had an 82 percent Arrive on Green (AoG) score. Metrics such as this could be used for the development of data-driven policy. The dashboard charts also showed vehicle density for when the signal was about to turn red—the timing of which could be adjusted to lighten the number of vehicles queuing.

Screenshot image of a white website with a blue graph, showing dark blue and red squiggles, which are traffic flow data at the intersection throughout the dat. Above, Dr. Tom Brennan can be seen explaining.

A live demo of the converted ATSPM dashboard demonstrated how useful the technology will be for making intersections more efficient.

It was clear that the conversion of ATSC data to ATSPM dramatically expanded the potential of every intersection in which it is equipped. The dashboard could be used to model changes in traffic flow, such as if a road diet were implemented, or if traffic from a major highway was diverted through the intersection. Safety benefits include data on red light violations that can be tabulated and used as justification for future improvements. One day data from connected vehicles could be integrated, too.

At the end of the presentation, Dr. Jin summarized their work: the team had innovated in converting raw data to ASTPM protocols that could then be used to boost signal performance and traffic flow optimization. This translation method avoids the intensive infrastructure cost of upgrading signals to ASTPM standards, saving money. An in-development model using Stop-Bar data will soon allow for real-time signal adjustment, letting traffic engineers tweak signal timings for optimal flow. At NJDOT, they are in the final stage of deploying this technology permanently on an agency server for future widespread use.

At the end of the event, several attendees asked questions through the platform’s chat feature.

Q. Inrix data does not provide individual probe data, how is it accounted for in the results?
Dr. Brennan: We’re not able to get individual vehicles, but it aggregates vehicle speed within one-minute increments. This then feeds that into ASTPM as if it were a single detector. Everything is within a confidence interval of 85 percent. The beauty of this software is that as long as you convert your information into the right format, you can put it in there.

Q. How do you control and change the cycle length tool?
Mr. McVeigh: Part of the adaptive system algorithm is to update cycle lengths in real time, based on the data being received. We can also provide guidance to the system on thresholds, on minimum and maximum lengths for cycles throughout the day. This is all for adaptive systems—coordinated systems use modeling to update their lengths. The tool used primarily at NJDOT is Synchro.

Q. How did the COVID-19 pandemic affect the data collection on Route 1?
Dr. Brennan: Because the researchers were working to calibrate a tool, the volume of traffic on the roadway did not affect their work.

Q. What are the biggest obstacles that you are facing in advancing this innovation around the state?
Mr. McVeigh: The first obstacle is ensuring that various datasets can be interfaced properly, because the ATSC system is providing data that is not necessarily compatible with ASTPM functions.
Dr. Brennan: There are also issues with code syntax, such as when SCATS is updated and logs data differently. Thanks to the graphical nature of the project, it is easy to see when this is happening.

Q. Did you face any issues in reconciling the Google Maps data with the CCTV data?
Dr. Jin: I think it was more with the video conversion. Google Maps provided good distance information that was then converted to become compatible with the video data. The critical step was coordinating these pixel coordinates to actual coordinates.

Q. What percentage of the adaptive signals are implemented on state highways?
Mr. McVeigh: Right now we have 118 adaptive signals in operation. Out of almost 2,600 signals in the state, a little under 20 percent of signals in the state are equipped with this technology.

Q. With multiple data sets fed into the system, how does it filter to avoid repetitions or duplicates?
Dr. Jin: We do have to filter the data that is fed in, and also have developed the logic that shows which line confirms the event occurred, and which line shows the starting point of the event. This is part of the translator work. In terms of different data sources, we were able to coordinate pretty well.

Q. Do you have any suggestions based on your research to help county and local governments advance the implementation of ASTPMs?
Mr. McVeigh: It’s a very powerful tool, make sure you have the practice to enable it to be used properly. From a technical standpoint it’s relatively straightforward, but the big thing is knowing how you want to use it—could be really effectively used as an empirical optimization tool. It all depends on the agency’s ability to do that.
Dr Jin: It is important to start knowing what is currently available, whether there is new construction or existing controllers, to see whether can deploy original ASTPM or these adaptive measures.
Dr. Brennan: It’s important to have strong IT support for these conversion activities. It’s not impossible, but necessary to have the support in place.

A recording of the webinar is available here, (or see right).

Resources

Federal Highway Administration. Automated Traffic Signal Performance Measures. https://ops.fhwa.dot.gov/arterial_mgmt/performance_measures.html

NJDOT Tech Transfer. (2018, December). What is an Automated Traffic Signal Performance Measure (ATSPM)? https://www.njdottechtransfer.net/automated-traffic-signal-performance-measures/

NJDOT Tech Transfer. (2020, June 12). Development of Real-Time Traffic Signal Performance Measurement System. https://www.njdottechtransfer.net/2020/06/12/development-of-rttspms/

Development of Real-Time Traffic Signal Performance Measurement System

Adaptive Signal Control Technology (ASCT) is a smart traffic signal technology that adjusts timing of traffic signals to accommodate changing traffic patterns and reduce congestion. NJDOT recently deployed this technology in select corridors and required a set of metrics to gauge functionality and effectiveness in easing traffic congestion and reliability. However, the monitoring and assessment of the ASCT performance at arterial corridors has been a time-consuming process.

The Automated Traffic Signal Performance Measures system (ATSPMs) developed by Utah DOT is one of the widely-used platforms for traffic signal performance monitoring with a large suite of performance metrics. One limitation of the existing ATSPM platform is its dependency on high-resolution controllers and the need to set up hardware and software at each individual intersection. Upgrading the existing controllers and reconfiguring the hardware and software at each intersection requires significant investment of funding and labor hours.

Recently completed research funded by the New Jersey Department of Transportation’s Bureau of Research mobilized researchers from Rutgers University, The College of New Jersey (TCNJ), and Rowan University to assist in advancing the goal of establishing automated traffic signal performance measures. The goals of the needed research were to develop a prototype Automated Traffic Signal Performance Measures platform for ASCT systems. The main focus was how to take advantage of the centrally-stored signal event and detector data of ASCT systems to generate the ATSPM performance metrics without intersection-level hardware or software deployment.

The study’s primary objectives were to examine: 1) how to utilize existing field data and equipment to establish Signal Performance Measures (SPMs) for real-time monitoring; and 2) identify what additional data and equipment may be employed to generate additional SPMs while automating the real-time traffic signal monitoring process. This research is especially important for New Jersey (NJ) with the deployment of ATSPM and the establishment of NJDOT’s Arterial Management Center (AMC).

Background

At present, NJDOT maintains a traffic signal system comprised of many types of equipment that affect signal performance, including different signal configurations and vehicle detection devices. Older equipment and ineffective detection technologies make real-time traffic signal monitoring quite difficult to implement across the state. With the implementation of more centrally-controlled traffic signal systems and the Department’s Arterial Management Center (the central control for remotely monitoring these signals) coming online, NJDOT needed standards to assure that the signals would operate properly and ease traffic congestion, and that the signals could be monitored remotely in real-time effectively.

ATSPMs are promoted by FHWA (Federal Highway Administration) as an EDC-4 (Every Day Counts 4) initiative. The use of ATSPMs has important foreseeable benefits:

  • Increased Safety. A shift to proactive operations and maintenance practices can improve safety by reducing the traffic congestion that results from poor and outdated signal timing.
  • Targeted Maintenance. ATSPMs provide the actionable information needed to deliver high-quality service to customers, with significant cost savings to agencies.
  • Improved Operations. Active monitoring of signalized intersection performance lets agencies address problems before they become complaints.
  • Improved Traffic Signal Timing and Optimization Policies. Agencies are able to adjust traffic signal timing parameters based on quantitative data without requiring a robust data collection and modeling process.
Research Approach

The research team recognized that the deployment of various adaptive traffic control systems such as InSync and SCATS systems on major NJ corridors and networks improved the capability for building real-time performance measures. The study included: a review of the literature and best practices; several stakeholder meetings; and recommendations and development of performance metrics, system architectures, data management, and strategies for deploying ATSPM systems using existing and planned NJDOT arterial infrastructure and technologies.

Figure 1: An Example real-time performance monitoring on County Road 541 and Irwick Road, Burlington County, NJ

Figure 1. An Example real-time performance monitoring on County Road 541 and Irwick Road, Burlington County, NJ

The researchers first conducted a literature review to identify examples of existing Signal Performance Measurement (SPM) systems to help inform the development of ATSPMs. The researchers described several exemplary initiatives, including the following:

  • In 2013, the Utah Department of Transportation’s (UDOT) SPM Platform was named an American Association of State Highway and Transportation Officials (AASHTO) Innovation Initiative. Deployed across the state, the system allows UDOT to monitor and manage signal operations for all signals maintained by the agency while aiding in more efficient travel flows along corridors.
  • From 2006 to 2013, the Indiana Department of Transportation (INDOT), with Purdue University, established a testbed of signal performance measures. INDOT developed a common platform for collecting real-time signal data, which became the foundation for AASHTO’s Innovation Initiative on Signal Performance Measures. This performance system has now been deployed at more than 3,000 intersections across the country.
  • Researchers at The College of New Jersey have established a signal performance measurements testbed using Burlington County’s centralized traffic signal management system. Traffic signal data collected along County Route 541 has been used to generate real-time performance measures and identify infrastructure improvements that could advance NJDOT’s ability to use real-time SPMs. An example of the existing real-time performance monitoring for Irwick Road and CR541 in Westhampton, NJ in Burlington County is shown in Figure 1.
  • Many state or local agencies including Pennsylvania DOT, Michigan DOT, New Jersey DOT, Lake County (Illinois), and Maricopa County (Arizona), etc., are actively incorporating ATSPMs into their traffic management and operation strategies. Lessons learned from implementation of ATSPMs from different agencies revealed that ATSPMs are critical to ATCS.

The research team organized and facilitated targeted stakeholder meetings. These meetings confirmed that stakeholders were not currently able to perform efficient real-time post-processing of the existing available data.  Through the meetings, the research team was able to scope more deeply into the type of performance measurements that were feasible and what could be done with the collected information.  Stakeholders also conveyed that the total number of operating adaptive signal intersections would more than double in the near-term future, making the need to efficiently process and leverage data from adaptive systems a more pressing concern. The discussions further confirmed that the big question for study was how to best leverage these adaptive systems to evaluate and manage future corridors.

Figure 2. Corridors where NJDOT has deployed ASCT systems; red denotes full operation, yellow denotes under construction, and blue denotes concept development

Figure 2. Corridors where NJDOT has deployed ASCT systems; red denotes full operation, yellow denotes under construction, and blue denotes concept development

The research team sought to better understand the inventory of NJDOT’s existing and planned ASCT systems. In 2019, New Jersey had over 2,500 NJDOT-maintained signals, but only 76 signals were on Adaptive Traffic Signal Systems.  In addition to the existing five corridors and the district in which ASCT systems had been deployed, 3 corridors were under construction and/or in final design and another 11 corridors were in the concept development phase for future ASCT installation at the time of the study (see Figure 2).

The research team visited the state’s Arterial Management Center (AMC) and investigated several signal performance systems – specifically, the Sydney Coordinated Adaptive Traffic System (SCATS), Rhythm Engineering’s InSync, and the Transportation Operations Coordinating Committee’s (TRANSCOM) real-time data feed – to better understand their interfaces, different types of detectors and their availability.

Figure 3. System Operation Data Flow Diagram

Figure 3. System Operation Data Flow Diagram

The research team designed an automated traffic signal performance measurement system (ATSPM) based on existing ATSPM open-source software to develop an economically justifiable ATSPM for arterial traffic management in New Jersey.  The entire system operates as shown in Figure 3. The high-resolution controller belonging to existing infrastructure is connected to an AMC at each signalized intersection. The controller event log file contains signal state data that is sent to an AMC database. The research team’s program automatically retrieves these data logs and translates the unprocessed data into a standard event code. The converted event file is inserted into an ATSPM database and the ATSPM software can generate signal performance metrics and produce visualizations to support performance-based maintenance and operations by traffic engineers.

Key Research and Implementation Activities

The research team successfully created a bench test of the ATSPM system based on data collected from high-resolution data from adaptive signal control systems including 13 SCATS locations on NJ Route 18 and 2 InSync locations on US Route 1. As a result of the testing, the research team successfully assembled a prototype for automated traffic signal performance measures in New Jersey.

1
2
10
3
7
6
4
8
9
5
previous arrow
next arrow
1
2
10
3
7
6
4
8
9
5
previous arrow
next arrow

Key research activities from the project are as follows:

  • Create Inventory of Existing NJDOT Arterial Management System: The team investigated several signal performance systems including InSync, SCATS, and TRANSCOM fusion application interfaces and different types of detectors and their availability. The team also conducted intensive review of state-of-the-art-and-practice of ATSPM system and identified ways of migrating the system to NJ.
  • Identify Performance Metrics and Measurement Methods for NJDOT ATSPM System: The team conducted a comprehensive review of SPMs built into an ATSPM system. The team investigated and customized SPMs that can be generated by NJDOT detector and travel time data.
  • Develop System Architecture and Concept of Operations for NJDOT ATSPM System and Established a Bench Test of ATSPM Located on TCNJ’s Campus: To leverage the existing ATCS system, the team developed a signal event conversion program to translate existing SCATS and InSync history log file to an event code that can be recognized by ATSPM. The detailed metrics are summarized in the figures to the right.
  • Prepare Real-Time Traffic Signal Data Management Guidelines: The research team created data management guidelines and a manual for data processing. The team validated the outputs through a comprehensive process. The team also completed a test to automatically connect to an ATSPM database using a VPN and MSSQL database management system.
  • Develop Deployment Strategies Considering Existing, Planned, and Future Systems/ Conduct Case Studies of System Deployment: The team initiated the pulling of one-month of data into their platform for the ATSPM. Large scale deployment of this system was expected to be conducted as part of Phase II research.

The research team observed that ATSPMs have distinct advantages over traditional traffic signal monitoring and the accompanying management process. The systems help shorten feedback loops with easier data collection and signal performance comparisons to enable before and after timing adjustments.

Future Work

In the first phase of the research project, the research team developed a software toolbox, NJDOT ATSPM 1.0.  The toolbox can convert the event output data from SCATS and InSync ATSC Systems into event data that can be processed by the ATSPM platform. The primary accomplishment was to integrate ATSPMs with existing ATCS from the centralized management console, instead of configuring at each controlled intersection on field. The proposed system bridges the gap between increasingly deployed ATSC and emerging ATSPMs without investment on new controllers. The effect of this research was validated on two selected corridors. NJDOT arterial management operators are able to use the ATSPM platform to generate key performance metrics and conduct system analysis for NJDOT’s ATSC corridors.

While the initial deployment and analysis was successful, it was limited in its scope. Phase II of the research involves the development and deployment of a significantly-enhanced version of the original toolbox, NJDOT ATSPM 2.0, along with a pilot study on the integration of ATSC controllers with Connected Autonomous Vehicle (CAV) technologies.

The research team will work with NJDOT to identify and add new performance metrics to generate additional Signal Performance Measures. The team can incorporate proprietary data from traveler information providers (e.g. INRIX and HERE) to generate other performance metrics such as queue/wait time, degree of saturation, predicted volumes, etc., and incorporate them into the NJDOT ATSPM platform. The team will also conduct pilot testing on the integration of Connected and Automated Vehicle (CAV), Roadside Units (RSU), On Board Unit (OBU) with the existing and planned NJDOT ATSC systems.

This developed ATSPM system from Phase II will bridge the gap between collected traffic data (e.g., signal controller data, detector data, and historical data) and needed performance information for decision-making. Phase II research is underway with an expected completion by November 2021.

Relationship to Strategic Goals

The development of RT-SPMs and the adapting and deployment of ATSPM with existing NJ ATSC systems is aligned with the FHWA EDC (Every Day Counts) Initiative to promote the rapid deployment of proven innovations. NJDOT ATSPM 2.0 will help meet the strategic EDC goal to accelerate the deployment of ATSPMs on existing and planned arterial corridors to reduce crashes, injuries, and fatalities, optimize mobility and enhance the quality of life.

The Phase II research supports the state initiative on advancing policy and testing of CAV technologies in New Jersey. The outcome of the project will be reported to NJDOT which is part of the New Jersey Advanced Autonomous Vehicle Task Force to make recommendations on laws, regulations and guidance to safely integrate advanced autonomous vehicle testing on the State’s highways, streets, and roads.


Resources

McVeigh, Kelly. (2019). Automated Traffic Signal Performance Measures.  Presentation at NJ STIC May 7th, 2019 Meeting.

Jin, P. J., Zhang, T., Brennan Jr, T. M., & Jalayer, M. (2019). Real-Time Signal Performance Measurement (RT-SPM) (No. FHWA NJ-2019-002).  Retrieved at: https://www.njdottechtransfer.net/wp-content/uploads/2020/01/FHWA-NJ-2019-002.pdf

Jin, P. J., Zhang, T., Brennan Jr, T. M., & Jalayer, M. (2019). Real-Time Signal Performance Measurement (RT-SPM) – Technical Brief Retrieved at: https://www.njdottechtransfer.net/wp-content/uploads/2020/01/FHWA-NJ-2019-002-TBrev.pdf

Zhang T., Jin P., Brennan, T., McVeigh, K. and Jalayer, M, Automating the Traffic Signal Performance Measures for Adaptive Traffic Signal Control System. ITS World Congress. 2020.

The Impact of SJTPO’s Traffic Signal Inventory on Signal Operations

As technology advances, so does the need for data—information that allows engineers, planners, and others to utilize innovative ways to improve transportation and safety. To implement smart traffic systems, whereby centrally controlled traffic signals and sensors regulate the flow of traffic, agencies must know the present state of their traffic signal infrastructure. The South Jersey Transportation Planning Organization (SJTPO), the metropolitan planning organization for four counties in South Jersey, sought to better understand their infrastructure by developing a database of all traffic signals in the region. Completed in 2017, the database provides local agencies with the information needed to target intersections and signals for upgrades and replacements. Replacement with newer integrated traffic signals improves traffic flow, allows for remote signal monitoring and regional signal maintenance, and supports bicycle and pedestrian improvements at intersections.

A traffic signal located in SJTPO’s region. (Source: Tracy, 2017)

In 2016, SJTPO sought to create a database for all traffic signals within Atlantic, Cape May, and Salem Counties. Previously, Cumberland County had developed a traffic signal inventory which SJTPO plans to integrate into the new, comprehensive database. SJTPO and county governments wanted to know the count, age, and types of signals in their jurisdictions. An SJTPO study in Vineland found that many of their signals were very old, with one using circa 1955 electromechanical components to operate. In addition, traffic signal maintenance progressively transferred from municipalities to counties and records of some signals were found to be deficient. The lack of information needed to properly maintain signals was a major impetus for creating the database, according to Andrew Tracy formerly of SJTPO (Source: Tracy, 2017).

Agencies across the country have created similar traffic signal databases. The Chicago Metropolitan Agency for Planning (CMAP), the regional metropolitan planning organization for Chicago and the surrounding seven counties, undertook development of a signal database in 2013 for the region, with the first version released to the public in 2018. CMAP’s goals for the database reflect those of SJTPO. The agency seeks to use the information for planning, and targeting specific signals and intersections for upgrades and replacement.

For an RFP issued to support its regional signal timing initiative,  SJTPO included a list of specific intersections identified by the counties for possible improvements. Extensive outreach to counties and municipalities to acquire signal data and plans took place prior to the database assembly to minimize the field work needed. For all data acquisition requiring field work, the subcontractor created an application to minimize errors with data input. The participating counties gave data collectors the keys to their controller cabinets along with a permission note in case police questioned them during their field work efforts. The signals were classified by features such as signal location, mast arm, head, sign, and presence of pedestrian push buttons. Additional information collected included intersection features such as ADA ramps, crosswalks, etc.

A look at SJTPO’s map and reviewer application for data input. (Source: Tracy, 2017)

Traffic data was also collected at identified intersections, including turning movement counts, queue lengths, delays, and travel times. This information could be used for traffic simulation modeling, performance measurement of intersections, and  revised signal timing plans. Extensive photography of the signals and intersections complemented the data set and provided visual aids. In total, 431 signals, including 258 traditional traffic signals and 173 beacons, were logged in the database across the 3 counties. The signal inventory was completed in 2017 and each county updates the database when a signal or intersection receives upgrades.

The traffic signal inventory database has created a variety of benefits for SJTPO and the region’s residents. One of the most noticeable benefits for local agencies has been access to data to target specific signals for upgraded technology, such as vehicle detection cameras and GPS clocks for signal coordination, or installation of new signals. The database can help identify intersections for bicycle and pedestrian facility improvements and greater accessibility for individuals with disabilities, such as wheelchair ramps and improved crosswalks. Signal upgrades benefit residents by improving traffic flow, and allowing for implementation of remote signal monitoring and signal maintenance at a regional, rather than local, level. Finally, the database reinforces knowledge preservation to ease any transitions in the event of staff turnover.

For other agencies considering a similar database, a Signal Inventory configuration is available via Collector for ArcGIS and performs similar functions as the SJTPO in-house application. Additional information on the process for assembling the SJTPO’s Traffic Signal Inventory Database can be found in a webinar (see below)  hosted by the Mid-Atlantic Geospatial Transportation Users Group.

Sources:

Chicago Metropolitan Agency for Planning. “Highway Traffic Signal Inventory: Draft Proposal.” CMAP, October 29, 2015. https://www.cmap.illinois.gov/documents/10180/481346/RegionnalSignals_Proposal_20151029_forRTOC.pdf/3aef6a03-a792-44ed-9515-11496c9c25f8.

South Jersey Transportation Planning Organization. “Request for Proposals: Regional Signal Timing Initiative.” SJTPO, July 13, 2017. https://www.sjtpo.org/wp-content/uploads/2017/03/SJTPO-RFP-Regional_Signal_Timing_Initiative.pdf.

Tracy, Andrew. October 30, 2017. The South Jersey Regional Traffic Signal Improvement Program. Presentation. https://www.sjtpo.org/wp-content/uploads/2017/11/CAC-10-30-2017-Andrew-Tracy-Signals.pdf.

Tracy, Andrew, Colleen Richwald, David Braig, and Matthew Duffy. October 12, 2017. https://www.youtube.com/watch?v=mMO7-NYuXZ0.

Getting through the Green: Smarter Traffic Management with Adaptive Signal Control

NJDOT Assistant Commissioner for Transportation Systems Management, C. William Kingsland, spoke about Adaptive Signal Control (ASCT) during the third Lunchtime Tech Talk hosted by the Bureau of Research on November 29, 2017.

The Federal Highway Administration (FHWA) defines ASCT as technologies that capture current traffic demand data to adjust traffic signal timing to optimize flow in coordinated traffic signal systems.  FHWA established ASCT as one of its Every Day Counts Round One initiatives in 2011-2012. New Jersey has implemented ASCT through the work of the Traffic Management Systems unit.

Assistant Commissioner Kingsland pointed out that commuters anticipate the time it will take for their typical commute routine and that reliability in travel time is important; people do not like fluctuation in the time it takes to get from A to B. When there is reliability of travel time, people’s expectations are met. ASCT effectively reduces congestion and fuel consumption, thus reducing complaints and frustration.

The ASCT system continuously learns based upon the traffic that is out there and will respond to changes in traffic patterns. Thus, the ability to adapt to unexpected changes in traffic conditions will produce improved mobility through a given area. Furthermore, as connected vehicles become more prominent, the system has the ability to gather information through Vehicle-to-Infrastructure communication and provide timely data of vehicle spacing and signal timing.

Assistant Commissioner Kingsland also provided some highlights about COAST- NJ, the management system developed by AECOM and the New Jersey Institute of Technology that is used to help decide where the ASCT systems will be placed. Using quantitative analysis, the tool ranks sections of corridors based on severity of congestion, variability of congestion, signal spacing, and traffic volume. COAST -NJ provides a classification system scoring process that encompasses 2,562 signalized intersections, 297 signalized arterial corridors, and 56 signal systems. It was officially released for use in March 2017.

During the Q&A portion of the Tech Talk, a member of the audience asked whether the system retains the collected traffic flow information to be able to look back to a certain date and time. The answer is that yes, it can. The issue, however, becomes length of records retention and where to store all of this information over the long-term.

In NJ, some of the NJDOT project locations with ASCT are along Route 130 (MP 69.79 to 74.51) with 15 intersections tied in; Route 168 (MP 6.79 to 9.72) with 11 intersections; and Route 32 (MP 0.0 to 1.20) with two intersections. Mr. Kingsland noted that Route 18 South in New Brunswick to East Brunswick is about to go online

Other agencies are also implementing ASCT. While not a NJDOT project, in the Meadowlands area there are 140 intersections tied into one ASCT system area managed by the Meadowlands Commission.

Mr. Kingsland was asked if rural areas with large distance between signals could possibly have cameras placed at intermediate sections between intersections. Kingsland replied that they certainly could, but the cost of such projects is prohibitive at this point in time.

Due to popular demand, Assistant Commissioner Kingsland presented this Tech Talk again on January 29, 2018.

Resources

Kingsland, W. (2017). Adaptive Signal Control—Getting Through The Green (Presentation).