Task Order 5404
Transit Operations Research


Toward Deployment of Adaptive Transit
Signal Priority (ATSP) System

Wei-Bin Zhang
California PATH

Summary

Transit Signal Priority (TSP) has been identified as a critical technology for deployment of Bus rapid Transit (BRT) systems and for improvement of traditional transit services. Applications of TSP for BRT and conventional transit applications have demonstrated the effectiveness in reducing total travel time and positive effects on improvement of the quality of transit service and increased ridership. Although the advantages is obvious, the deployment of TSP has been slow and mostly limited to selected corridors primarily due to two factors, including (1) the concerns from traffic operation authorities on potential impact to other traffic caused by frequent priority interruptions and (2) the capital cost for instrumenting the signal priority system.

California PATH Program, in partnership with Caltrans and San Mateo Transit Authority (Samtrans), are conducting a research program to investigate an Adaptive Transit Signal Priority (ATSP) concept. The PATH ATSP concept uses real-time GPS position, bus movement information and historical bus travel behavior data to predict the bus arrival time to the intersection. The predicted bus arrival information provides much longer lead time to allow the traffic controller to be 'adaptive' to the bus arrival as well as the traffic situations and make it possible to distribute the impact to several control cycles.

Objectives

(1) investigate methodologies for integrating the PATH APST system concept with the existing GPS/ACS systems that will allow bus positions to be sampled on a needed basis, and

(2) further develop ATSP algorithm to work with more advanced traffic control devices (such as 2070) so that adequate signal priority can be given when needed and impact to other traffic will be minimized.

The results of this study will include a dynamic polling algorithm that would allow the existing GPS/ACS to support ATSP data needs and form recommendations to the transit requirements for TSP and GPS/ACS requirements and standards that will facilitate the integrated application of GPS/ACS and signal priority technologies. The results will also include improved ATSP algorithm. These results are of immediate interests to transit agencies such as Samtrans for deployment of cost-effective TSP.

Methodologies

The proposed study can be accomplished by adopting the following methodologies:

Objective One

In order to address the first objective defined in this study, methodologies for GPS/ACS system to provide timely priority request must be addressed.

Two approaches can potentially enable the existing GPS/ACS to provide timely priority calls.

The first approach is to develop a Dynamic Polling Algorithm (DPA) to enable buses to be only sampled on a needed basis.

In order to develop the DPA algorithm, the following investigation will be conducted:

  • Study of transit management data needs: We plan to work with Samtrans and Orbital Co. to study the characteristics of bus data and transit data needs in order to define the minimum requirements for flexible data polling for accommodating both transit management and signal priority needs.
  • Develop DPA strategies: Polling strategies, likely to be location and event dependent, will be developed.
  • Conduct simulation: Simulation of ACS using DPA strategies on a scaled Samtrans service network and fleet will be conducted.

The second approach is to develop the 'Priority Polling' algorithm (PPA) specifically to take advantage of the additional bandwidth for priority vehicles.

The study of this approach will involve:
  • Analyze ACS data from Samtrans to understand bus travel behaviors. Through this analysis, we'll determine statistically the maximum number of buses that could arrive at intersections at a given time interval and criteria for bus to receive priority status.
  • Develop an priority PPA algorithm that will be able to determine which bus needs to receive priority and the polling protocols.
  • Conduct simulation: Simulation of ACS using PPA strategies on a scaled Samtrans service network ad fleet will be conducted.
We will use the data to conduct simulation studies of the both approaches using Paramics for the El Camino Real. Trough the simulation, we will evaluate the viability of both algorithms and to improve the one(s) that proven to be feasible. Field testing will be performed using the three Samtrans testing buses to demonstrate the effectiveness of the selected approach(es).

Objective Two

In order to allow the signal priority system to better accommodate bus signal priority calls while minimizing the impact to the other traffic and maintaining the coordination among signals, a signal priority algorithm that is based adjustable cycle length signal control scheme must be explored.

In order to allow the signal priority system to better accommodate bus signal priority calls while minimizing the impact to the other traffic and maintaining the coordination among signals, a signal priority algorithm that is based adjustable cycle length signal control scheme must be explored.

The proposed research will focus on following issues:
  • Model flow dynamics representing queue formation and oscillation between signalized intersections by using loop detector data
  • Optimize cycle length and phase duration at each intersection based on the dynamic platoon model developed in phase one with considering bus priority requests
  • Conduct simulation studies using microscopic simulation model to test and validate the two models, make MOE (measurement of effectiveness) analysis
For the quasi-adaptive operation we will look at:
  • Design of more detailed signal coordination plans offline based on historical traffic data
  • Modification of current signal systems so that they can switch between different coordination plans more constantly based on system detector data and bus priority calls
  • Simulation studies to test and evaluate the coordination plans

Scope of Work

Task 1. Develop Approaches for Bus Position Sampling and Priority Calls
Task 2. Simulation Analysis of Bus Position Sampling and Priority Calls Approaches
Task 3. Conduct Field Testing of ATSP using emulated GPS/ACS System
Task 4. Further Develop ATSP Algorithm
Task 5. Simulation Study of the Enhanced ATSP Algorithm
Task 6. Field Testing of the Enhanced ATSP Algorithm