Deriving Signal Performance Metrics from Large-Scale Connected Vehicle System Deployment

This paper documents the development of signal performance metrics (SPMs) from connected vehicle data, including application to existing deployment locations in the United States. The metrics are aggregated from anonymized vehicle traces traversing signalized intersections that are part of a system deployment that is completely based on existing communication and signal control infrastructure. No retrofit to controllers is necessary. The system structure consists of (1) traffic signal data collection via real-time data polling, (2) signal state prediction and Signal Phase and Timing (SPaT) message generation, (3) data dissemination of SPaT and Map data (MAP) messages, and (4) in-vehicle applications including countdown timers, speed advice to avoid stops, and emerging applications such as powertrain management or automatic engine start/stop functions. Four vehicle metrics were constructed including a velocity profile, arrivals by phase state (green, red), delay, and split failures. A large-scale case study in the City of Frisco, TX showed potential in helping daily management of traffic signal control, and potentially improving traffic flows. The connected vehicle SPMs were imported and visualized in a business intelligence tool (Microsoft Power BI) to deliver a signal intelligence report comprised of a series of interactive data dashboards. This interactive report provides a web-based or stand-alone interface to individual signals, or corridor or citywide measures of average vehicle delay, split failures, and arrival states.