In the United States, more than 300,000 traffic signals are currently in operation. According to the Federal Highway Administration, the operation and performance of most of these signals are assessed through citizen complaints. Historically, agencies have been forced to rely on manual counts input into software and simulation models to develop timings, with the presumption that if there are no public complaints, then everything is working acceptably, often compromising on performance and efficiency as user demand changes.
Automated Traffic Signal Performance Measures (ATSPMs) started in the mid-2000s with the collection and analysis of high-resolution event-based data for traffic signal performance. Since then, researchers at Purdue University along with practitioners at the Indiana Department of Transportation, Georgia Department of Transportation, and Utah Department of Transportation have evolved the use of event-based data into a method of assessing and improving the performance of traffic signals, traffic signal systems, and traffic signal system business practices. From a technical standpoint, the suite of ATSPMs can allow an agency to monitor capacity, progression, multimodal, and maintenance performance measures without the added expense of a central or adaptive traffic signal system. These performance measures can be developed through robust communication and typical traffic signal detector information, though additional detection is required to take advantage of all the performance measures. For the sake of consistency, introduction of commonly used terminology throughout this proposal include high-resolution controller data schema, geo-spatial metadata, and signal timing plans.
The experiences of users of ATSPMs and those who have been actively developing the methodology have revealed the following limitations of the current high-resolution data scheme on which the performance measures are founded:
· Lack of common methods for describing geo-spatial information (“metadata”).
· High resolution controller data only stores changes in states.
· Unavailability of data management and archival process.
· Lack of holistic performance measurement.
The purpose of this research is to address shortcomings of the current conceptions of high-resolution data on which ATSPMs are founded. This will “future-proof” the data and the systems that rely on it by enabling it to integrate new data streams and enhance its scalability and transferability to multiple agencies and jurisdictions. The main objectives of this research should include:
· Reconcile the diverse approaches taken to the intersection metadata problem: for example, the conception of an intersection in SAE J945 standard as compared to methods of describing detector layouts for signal control.
· Develop a standard schema for saving geospatial information required for current ATSPM systems.
· Develop an open-source tool for easily generating the geospatial information for a given intersection.
· Expand dynamic data storage standards, schema, and format to include storage of short-term trajectories available via BSM message or advanced sensors.
· Design data compression and aggregations methods to enable long-term storage of critical information to manage economics and processing speed of the system.
· Enhance the performance measures to provide a more holistic view of the system.
The results of this research will be used by Traffic Operations and Traffic Signal Management of agencies. The tool and recommendation developed during the course of this research will make it easier for DOTs to implement a scalable system which will consume data from Connected Vehicles in the future. The tools could be immediately applied by agencies and vendors implementing the ATSPM system.