Peachtree Corners - one of the first US smart cities - has launched a pilot project with traffic and pedestrian safety solutions provider ConnVas.
The city, in the US state of Georgia, said the pilot project will address the safety challenges of busy crosswalks by testing how real-time AI-driven detection and alerts can reduce accidents. It works by activating alerts - even when pedestrians do not activate crosswalk signals.
At the core is the ConnVas Magic (Multi Access Generic Infrastructure Controller) detector. The edge computing system incorporates cameras mounted on RRFB (rectangular rapid flashing beacon) poles to monitor pedestrian and vehicle movements. It also has Vehicle to Everything (V2X) transceivers for real-time Vehicle to Infrastructure (V2I) communication and an integrated audio-visual alert system.
When the system detects a pedestrian or cyclist using the crosswalk, crossing outside of the crosswalk lines or a vehicle approaching the crosswalk, the technologies work together to immediately activate alerts to both the users of the crosswalk and incoming drivers, even if the person did not activate the crosswalk alert.
The system continually collects anonymous data for ongoing analysis of crosswalk usage and traffic.
The system has been set up just outside a school. “By integrating advanced AI detection and real-time alerts, this ConnVas pilot enhances safety in an area where our residents need it the most today,” said Brandon Branham, executive director of Curiosity Lab, a 5G-enabled intelligent mobility and smart city living laboratory near Peachtree Corners.
Preliminary results from the pilot programme that were extracted from a mid-week 72-hour period, show distinct patterns. ConnVas reported over 220 pedestrian crossings for more than 70 crossing events, with only three false alarms. The system consistently and successfully activated flashing beacons and audio alerts in real time, ensuring those crossing the road and oncoming drivers were promptly warned.
Additionally, traffic data collected by the system included felonies, near-misses and counts of pedestrians, cyclists and vehicles. The information is being used to evaluate the technology’s impact and will aid city officials to make informed decisions about improving road safety and urban planning in specific locations.
“Through continuous testing and data analysis, we’ve been able to improve the accuracy of our system’s AI detection, ensuring faster and more reliable activation of alerts,” said Danny Frydman, founder and chief executive of ConnVas.
“This has directly enhanced the safety of the area by addressing critical challenges including missed warnings for pedestrians and cyclists or delayed driver awareness. The insights gained from this pilot are helping us evolve the software to adapt to complex traffic scenarios and create safer, smarter intersections everywhere,” said Frydman. “We hope to scale this solution throughout the city as the pilot and technology continues to develop.”
ConnVas said that the pilot in Peachtree Corners is continuing to evaluate various features including autonomous operation features that eliminates the need for manual button presses and reduce false alarms, cyclist-friendly designs and V2X connectivity with connected vehicles to alert drivers of crosswalk activity.
Additional capabilities such as dynamic illumination based on pedestrian location are being explored to further enhance VRU - vulnerable road user - safety.
Peachtree Corners, a city of more than 45,000 residents, says it was the first to deploy teleoperated e-scooters and fully autonomous shuttles for use by residents. There is also a solar roadway and the largest electric vehicle fast-charging hub in the region.
Curiosity Lab was set up to be a proving ground for IoT, mobility and smart city emerging technologies. The centrepiece of the lab is a 5km-long autonomous vehicle lane leveraging cellular vehicle-to-everything (C-V2X) technologies. Additional infrastructure includes intelligent traffic cameras and traffic signals, smart streetlights, the country’s first “IoT Central Control Room” implemented in a city and a 25,000 square foot (2,320m²) technology incubator.
