AIWaysion Concludes Successful Curb Activity Pilot with NYC DOT

AIWaysion's Edge AI CurbVision system maintained 100% uptime and achieved more than 95% accuracy while continuously monitoring curb activity at three Upper West Side locations

AIWaysion recently completed a 15-month curb activity pilot with the New York City Department of Transportation (NYC DOT) as part of the Transit Tech Lab’s Curb Activity Challenge.

Deployed at three locations on Manhattan’s Upper West Side, AIWaysion’s Edge AI CurbVision system operated 24/7, processing curb occupancy and event data locally on-device without cloud-based video analysis.

Transit Tech Lab Pilot Summary: "AIWaysion deployed its Edge AI curbside camera system at three Upper West Side locations, monitoring six curb zones on both sides of the street for 15 months to automate curb activity data collection that NYC DOT previously captured through manual time-lapse video review. The system ran with 100% uptime across winter and nighttime conditions, detecting 500 to 800 curb events per day at over 95% accuracy on the camera-facing side, using edge processing with no fiber or cloud infrastructure required. AIWaysion built NYC-specific classifications including outdoor dining, cargo bikes, and waste management vehicles, and delivered a web dashboard giving DOT staff curb activity and historical analytics by time of day, vehicle type, and lane type. If scaled citywide using solar-powered, cellular-connected units, AIWaysion could give NYC DOT a continuous data layer to inform enforcement, capital planning, and curb regulation."

Continuous, On-Device Curb Monitoring

Beginning with the initial deployment on December 11, 2024, the CurbVision system maintained 100% uptime throughout the monitoring period, including winter and nighttime conditions.

Each CurbVision camera monitored curbs on both sides of the street. The system achieved more than 95% curb occupancy accuracy on both sides and more than 95% curb-event detection accuracy on the camera-facing side. Curb-event records included start and end times, unique event IDs, activity classifications, and affected lane types.

CurbVision detected and classified a range of curb activities, including:

• Parking and double parking

• Deliveries

• Micromobility activity

• Outdoor dining

• Waste management

• Emergency use

The system also identified whether an activity affected a parking lane, travel lane, buffer zone, or bike lane. Its data structure followed the Open Mobility Foundation’s Curb Data Specification, supporting standardized use and integration of curb data.

A Curb Data API and web dashboard (WaysionNet Curb) provided NYC DOT staff with access to real-time monitoring, historical analytics, and operational insights by time of day, activity type, vehicle type, and lane type.

NYC DOT Feedback and Opportunities to Scale

As part of the project closeout, Dr. Wei Sun, Co-Founder and CEO of AIWaysion, and Brian Brooke, Executive Vice President of AIWaysion, presented the results and lessons learned at NYC DOT headquarters to transportation professionals representing multiple divisions.

NYC DOT project managers Jonathan Hawkins, Director of NYC Streets Plan, and Matthew Garcia, Deputy Director of Parking Planning, acknowledged the AIWaysion team’s work and the successful performance of the CurbVision system.

Hawkins and Garcia emphasized the value of continuous, automated curb data collection, real-time visibility, and actionable insights. They also highlighted the technology’s potential to support more automated decision-making for curb enforcement and transportation operations.

The discussion explored several opportunities to scale the system, including optimizing camera placement, using solar-powered deployments to reduce installation costs, and leveraging existing traffic cameras for curb monitoring. These approaches align closely with AIWaysion’s existing technical capabilities and deployment options.

“NYC DOT’s operational input was essential to adapting CurbVision to real-world curb conditions and NYC-specific needs,” said Sun. “As the new Office of Curb Management begins its work, we look forward to supporting its vision with scalable, fine-grained curb data for both real-time operations and long-term planning.”

Curb space is highly dynamic and contested and can sometimes create unsafe conditions for road users. Reliable, 24/7 curb activity data can help agencies better understand how curb space is used and support enforcement, operations, safety planning, capital investment, and curb regulation.

Project Acknowledgments

AIWaysion sincerely thanks Jonathan Hawkins, Matthew Garcia, Carl Sundstrom, and Noah Dubin of NYC DOT for implementing the pilot, supporting the CurbVision deployments, and providing valuable operational input—including the development of NYC-specific curb-event categories and classifications.

AIWaysion also thanks Stacey Matlen and Chloe Rosenberg of the Transit Tech Lab for leading and managing the program, and Yiqiao Li of the City University of New York for conducting the independent third-party evaluation.