KLK133: Improved Signalized Intersection Performance Measurement, Phase II

Principal Investigator

Project Objectives

1. Develop performance measurement methodologies for locally controlled actuated signalized intersection operations
2. Develop learning objects for students to understand operations of actuated signalized intersections at different levels

Task Descriptions

Task 1: Develop and test improved methods for detecting phase failure. Testing will be focused on the ability of a method to detect phase failure using standard detector configurations. Benchmark data were obtained manually from field video footage and simulation data were obtained from a large set of VISSIM/HILS experimental runs. Detection errors will be quantified on the basis of omission, commission, and overall error rates. Both field and simulation data will be used to complete the testing. Acceptable levels of phase failure will be recommended, as will acceptable levels of phase failure detection error. Automated methods will be implemented using Autoscope and the current set of field video recordings.

Task 2: Develop and test improved methods for measuring delay. Delay measurement methods will be restricted to information available, given standard detector configurations. Benchmark data were obtained from field video footage. Methods will be automated using Autoscope. Assumptions of the proposed methods will be thoroughly tested to determine their robustness to different types of errors and error levels

Task 3: Establish and test controller vs performance measurement relationships to improve intersection traffic operations. This task will focus on using the following measures for evaluating intersection performance and diagnosing problems, if they exist. 1. phase failure, 2. delay, 3. green-time utilization, 4. detector occupancy, 5. speed, 6. phase indication, 7. headway, and 8. count data.

These measures will be related in some systematic fashion to a limited set of controller parameters/settings. Problems associated with the following controller settings will likely be considered in this task. 1. passage time, 2. protected left-turn phasing, 3. detector delay, and 4. max green time.

The objective of this step is to arrive at an understanding that would facilitate improving traffic operations, never knowing if optimal operations are actually achieved.

Task 4: Devise a measure to communicate the plausibility of operational improvement. While an intersection may operate poorly, it is possible that significant improvements may result from changing controller parameters/settings. Completion of this task will result in a measure that will describe this likelihood. This task will start with empirical measures, such as intersection saturation. However, it is likely that any one of these measures would need to be augmented by others, in order to provide a comprehensive picture of how flexible intersection operations are.

Task 5: Documentation and publication. • Final Report • ASCE paper (Tung): Automated delay measurement w/ study of error effects. • TRB paper (Wu and Azizur): Local intersection performance measurement systems approach. • TRB paper (Wu and Madrigal): Using phase failure and delay information to diagnose and improve actuated control.

Task 6: Develop learning objects for students at senior level. Learning objects for students at the senior level will be created via CE 474, where new objects will be created for the course while it is being taught during the 2006 fall semester. Emphasis of the training materials will be on the evaluation of actuated signalized intersection performance.

The learning objectives of this learning material are as follows:

• Explain basic signal control operations of isolated actuated signalized intersections (vehicle detector calls, passage time, minimum green, maximum green),
• Assess the quality traffic signal operations in terms of green utilization, responsiveness, and delay, and
• Perform signal timing modifications for isolated intersections.

VISSIM or CORSIM microscopic simulation software, traffic control hardware, and field data will be used as needed to provide the best hands-on learning outcomes for the learning objects.

Milestones

1. November 2006--Develop measure to communicate the plausibility of operational improvements
2. January 2006--Develop and report improved method for measuring delay. Li-Wei’s paper submitted to ASCE
3. June 2007--Establish and test controller vs performance measurement relationships to improve intersection traffic operations
4. July 2007--Submit two papers authored by Wu, Madrigal, and Azizur to TRB
5. August 2007--Develop learning objects
6. September 2007--Submit final report

Budget Information

Student Involvement

Undergraduate student role: Student will assist in data analysis for Task 3 and Task 4. The students will also assist in beta testing learning materials as part of Task 6 [Academic year (total project expense for both students): 10 hours/week; summer (one student): 40 hours/week].

Graduate student role: Student will assist in Tasks 3, 4, and 6 and will assist in Task 5. (Academic year: 20 hours/week; summer: 30 hours/week).

Graduate student role: Student will lead in Task 3, 4 and 6 and will assist in Task 5 (Academic year: 20 hours/week; summer: 40 hours/week).

Graduate student role: Student will lead in Task 2 and will assist in Task 5 (fall semester: 20 hours/week)

Graduate student role: Student will lead in Task 1 and contribute to Task 3 (academic year: 20 hours/week; summer: 30 hours/week)

Relationship to the NIATT Strategic Plan and to Other Research Projects

Continuation of existing research project: This project is a continuation of a NIATT project entitled, “Development of New Actuated Signalized Intersection Performance Measurement Methodologies Using Traffic Controller Input and Output Data”. Conclusions of this research project indicated that research was needed in the following areas:

1) Performance measurement must be done with standard detector configurations. 2) We need to test and develop better phase failure measurement methods, because the current methods have not been tested until this project and these methods are inaccurate. 3) Methods for automated delay measurement with standard detector configurations need to be tested more thoroughly and improved. 4) More research is needed to understand how to integrate performance measurement with the task of improving traffic operations.

Supports Center of Traffic Operations and Control mission: Research activities of this project are consistent with the mission of the Center for Traffic Operations and Control, which “is to develop, test, and evaluate technology products supporting advanced traffic control systems” . This project also involves three of the five components identified as the focus of this center, which are traffic detection, control, and simulation.

Supports NIATT Strategic Plan: The proposed project will produce performance measurement methodologies for actuated signalized intersections. Locally controlled actuated signalized intersection operations will improve as a result of the implementation of products from this project. Hence, this project supports one of the three federal or state priorities as identified in NIATT’s strategic plan, which is “Traffic control system technologies that are essential to the national intelligent transportation infrastructure.”

Supports NIATT/UTC program goals: This project supports the NIATT/UTC program goals in the following four ways:

• Human resources: One civil engineering faculty, two undergraduate students  and four graduate students will be involved in this project.
• Education goal: Learning materials developed in the proposed project will furnish opportunities for students to develop their skills for evaluating actuated signalized intersection performance in NIATT’s lab, where the tools are available: workstations, simulation software, and traffic control devices.
• Research performance goal: Researchers will benefit from this project because of the resulting access to a time series of traffic detection and controller response data, which enable detailed phase-by-phase analysis, leading to improved understanding of controller operations.
• Technology transfer goal: Products from the proposed project will be used in two ways. The research team will write reports and peer reviewed articles to discuss and disseminate the results. Learning materials will supplement course materials in intersection operations.

This project is closely related with other research projects in the following ways:

This project will continue to support the MOST project’s vision of providing a realistic simulation environment of operations for actuated traffic controllers and improving students’ understanding of “the interaction of various design and operation parameters of the traffic signal control system”

The proposed project supports the MOST project because of its focus on the evaluation of local actuated signalized intersection operations. In addition, the proposed project offers support through its use of simulation based and field data based testing procedures that can then be used in the MOST project.

The proposed project will use raw and extracted field data and simulation output data from the UTC Data Systems project as input to the testing procedures. No additional effort is needed to acquire the data because it is already available.

The NGSIM Video, obtained from Cambridge Systematics/FHWA, will be used to support the testing of automated methods for measuring delay.

Relation to FHWA Research Gaps and Opportunities

This project is very closely aligned with the following research gaps/opportunities presented at the FHWA/UTC workshop in Washington, D.C. on June 6-7, 2006:

• Research methods for identifying saturated conditions at signalized intersections and development of signal timing strategies to respond to this condition.
• Research methods for using existing traffic signal system infrastructure to assess system performance in real-time and evaluation of the impacts of traffic signal timing changes prior to implementation.
• Research methods for evaluating travel time on arterials routes and assessing the needs for and benefits of this information.

Technology Transfer Activities

Two technology transfer products will be generated from this project: 1) reports and publications discussing research activities and results; and 2) learning materials for university students. The papers that will be submitted as a result of research funded through this project are as follows: a. ASCE paper: Automated delay measurement w/ study of error effects. b. TRB paper: Local intersection performance measurement systems approach. c. TRB paper: Using phase failure and delay information to diagnose and improve actuated control.

Potential Benefits of the Project

New methodologies resulting from this project will add valuable tools to the area of performance evaluation for local actuated signalized intersections. The performance measures will allow traffic engineers to evaluate actuated signal timing plans using controller output instead of solely relying on traditional field data or field observations. While traditional field data are important they do not communicate information that fully describes the quality of actuated control (e.g., green utilization and premature phase termination) because they are usually aggregated over time. Also, these methods will benefit practice by improving the engineers’ ability to monitor, evaluate, and update signal systems. Furthermore, results from these methods will establish some relationships between signal controller parameters and intersection performance, which will strengthen our ability of how to evaluate and improve signalized intersection operations.

This project will further our position as a significant player in achieving the SHARP II research plan, which emphasizes travel time reliability. This is an important characteristic of this project, because the SHARP II research plan will play a role in the directions that NIATT proposes to take after the competition for Tier 1 funding.

Project status

Active

Final Report