MTA, Google partner on AI track maintenance project

Th Metropolitan Transportation Authority last week announced it is testing a track defect detection program in partnership with Google Public Sector.

The partnership will build on the success of Google’s existing TrackInspect prototype, MTA officials said in a press release.

The TrackInspect prototype program retrofitted Google Pixel smartphones onto R46 subway cars on the A Line to capture vibrations and sound patterns, MTA officials said. Data is sent to cloud-based systems, where it's analyzed by artificial intelligence and machine learning algorithms to create predictive insights. Track inspectors can then check the locations highlighted by the system.

“The TrackInspect pilot program identified 92% of the defect locations found by track inspectors, illustrating that enhanced data analysis can help expedite problem identification and resolution to improve railway reliability,” said Brent Mitchell, Google Public Sector's vice president, go-to-market.

In an initial pilot, TrackInspect collected 335 million sensor readings, 1 million GPS locations and 1,200 hours of audio. The information collected by the prototype complements the data provided by MTA’s track geometry cars, MTA officials said.

MTA has released a request for expressions of interest for other companies who have developed sensors or analytical capabilities.