Active Projects New York City Department of Transportation

New York City DOT launches AI for real-time roadway management and proactive maintenance

New York City Department of Transportation has been partnering with Blyncsy since June 2023 to run real-time, AI-powered crosswalk detection and paint line scoring and to develop a historical model to track road condition changes over time.


reduction in manual surveys


Potential savings using Blyncsy compared to manual or LiDAR inspections


pounds of carbon emissions saved per vehicle per year


New York




Department of Transportation

In this Article

Share ⟶

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.
John Doe

Solving Unique Challenges

The New York City Department of Transportation is a massive organization with an even larger responsibility; manage over 6,300 miles of roads, 800 bridges, one million street signs, 315,000 street lights and over 1,600 crosswalks. With a focus on pedestrian safety and enhancing the overall safety of the entire transportation network, NYCDOT partnered with Blyncsy to automatically inspect all 1,650 crosswalks and, using artificial intelligence models developed by Blyncsy, analyze the condition of the paint lines within each crosswalk.

Crosswalk Detection and Scoring

Through the partnership with New York City DOT, Blyncsy will conduct monthly assessments and detections for the following areas of interest on 500 intersections and approximately 1,650 total crosswalks:

  • Crosswalk location identification utilizing OSM for geographical information, incorporating NYCDOT intersection nodes and street linestrings
  • Utilizing footways to identify intersecting points for crosswalk polygons, which enhances the precision of crosswalk mapping and allows for comprehensive coverage of all crosswalk locations
  • Imagery captured from crowd-sourced dash cameras allows our models to detect crosswalks and provide a condition rating.
  • The detection box visually extracts the crosswalk from the image, processing algorithms determine the percentage of paint remaining in each crosswalk.
Crosswalk detection and automated scoring
New York City crosswalk detection with percentage of paint remaining

Success in New York City

Through the initial phases of this project, the following results were observed:

  • Crosswalk detection and scoring improved as the data set grew larger and additional crosswalks were added
  • Accuracy and reliability of scoring system in determining real-world visibility of crosswalks 
  • Opportunities were opened to adding additional cameras on city vehicles, including sanitation trucks and other maintenance vehicles.

Explore more Active Projects