THIS subway has installed smartphones on certain subway cars in hopes of gathering data to prevent track issues.
Artificial intelligence systems are being trained with that information — possibly taking over the jobs of transit workers.
2
2
Rob Sarno, MTA’s assistant chief track officer, has been trying to come up with creative and innovative solutions to fix common issues with the New York City’s subway system.
For years, Sarno has looked into how artificial intelligence could make his job and the millions of people who work for the transit system a little bit easier.
What he discovered was he could teach an AI system what a railroad sounds like when there’s something wrong with it.
He developed a pilot program between the MTA and Google Public Sector that would intake data collected from Google’s Pixel smartphones that were strapped to certain subway cars, as reported by CNN.
How it works is several smartphones were installed on four subway cars which collected 335 million sensor readings — a total of 1,200 hours of audio, according to the MTA.
The phones were placed inside and underneath the subway cars between September 2024 and January 2025.
Any sounds and vibrations that were picked up by the smartphones were then used to figure out where there might be a track issue.
The MTA assured the public that the smartphones did not intake any riders’ information or conversations.
The phones inside subway cars were only able to capture vibrations.
This data was then fed into an artificial intelligence model that would examine the sounds for any indication of a defect.
The program, called TrackInspect, uses sensor hardware with advanced cloud and artificial intelligence to identify track problems.
But it’s not all up to the artificial intelligence.
New York City Transit track inspectors confirm the issues the TrackInspect detects, providing continuous feedback to better train the model.
Sarno hopes that the new technology will help prevent MTA delays.
What is TrackInspect?
MTA partnered with the Rapid Innovation Team at Google Public Sector to create a prototype called TrackInspect that can detect potential track issues.
The prototype uses sensor hardware with advanced cloud and artificial intelligence to identify track problems.
New York City Transit track inspectors also double-check the issues the TrackInspect detects to confirm whether or not the prototype got it right. They provide continuous feedback to train the model.
The MTA hit one billion trips on the New York City subway in 2024, as announced by Governor Kathy Hochul.
“There’s no better way to get around New York City than taking the subway, and we’re proud to be setting post-pandemic ridership records while investing in major new initiatives that will modernize our transit system,” she said.
Despite millions of New Yorkers who depend on the subway system to get them around the city, the MTA still faces a number of delays on a daily basis.
According to MTA data, there were more than 38,000 total delays in September and 39,000 in October of last year.
With new initiatives like TrackInspect, the MTA hopes to reduce the interruptions and delays for city-goers.
The prototype has correctly identified over 90 percent of the defect locations found by the MTA’s inspectors, a Google Public Sector spokesperson told CNN.
