Sensys Networks is partnering with the University of California, Berkeley and Hyundai America Technical Center Inc. (HATCI) as part of the US government’s Advanced Research Projects Agency-Energy (ARPA-E) project to develop control systems for electric CAVs.
The partners are working on ARPA-E’s NEXTCAR (NEXT-Generation Energy Technologies for Connected and Automated On-Road Vehicles) program and are developing an innovative vehicle dynamics and power train (VD&PT) control architecture based on a predictive and data-driven approach, which will optimize plug-in hybrid electric vehicle (PHEV) performance in real-world conditions, and facilitate efficient departure at intersections, predictive cruise and speed profiles, and learning-based eco-routing and tuning.
As a leading provider of integrated wireless traffic detection and data systems, Sensys Networks is using predictive analytics tools to combine historical data with real-time data at signalized intersections. The company is then accurately estimating the remaining time in the green or red light phase and delivering it securely to connected and autonomous vehicles (CAVs). Sensys Networks says this particular project demonstrates its vision to make use of vehicle connectivity with the transportation infrastructure, and automation technologies to improve vehicle controls and powertrain operation.
The challenge in delivering accurate signal phase and timing (SPAT) data is the predictive nature of the problem. The length of the green or red phase depends on real-time traffic, as most traffic signals would extend the green phase as more cars approach the intersection. Sensys Networks has developed a breakthrough system that delivers Predictive SPAT data to CAVs. The solution uses the Sensys Networks FlexControl Edge Gateway and SNAPS architecture of its SensTraffic software platform to collect and store signal phase and vehicle detection data.
“An autonomous vehicle approaching a signalized intersection with accurate knowledge of whether the signal will be green or red when it gets there, will optimize its approach to minimize stops or accelerations/decelerations and can maximize its energy consumption,” explained Amine Haoui, Sensys Networks’ CEO.
“A connected vehicle stopped at a traffic light may switch its engine off if it knows that it will be 20 seconds until green, but may determine it’s more efficient not to switch off if it’s only three seconds to green.”
Professor Francesco Borrelli, of UC Berkeley and the principal investigator for the NEXTCAR project, said, “At Berkeley we are building the intelligence beyond the next generation of CAVs. This project with Sensys Networks demonstrates our vision to leverage vehicle connectivity with transportation infrastructure, and automation technologies to optimize vehicle operations.”