The Metropolitan Transportation Authority (MTA), in partnership with Google Public Sector, recently launched a pilot program that will detect potential track defects before they escalate into operational issues that disrupt service.
The program builds on the success of the TrackInspect prototype that collected 335 million sensor readings, 1 million GPS locations, and 1,200 hours of audio that was combined with New York City Transit’s (NYCT) database of track nonconformities. That data was analyzed by a machine-learning model running on Google Cloud.
“By being able to detect early defects in the rails, it saves not just money but also time – for both crew members and riders,” Demetrius Crichlow NYCT president, said. “This innovative program – which is the first of its kind – uses AI (artificial intelligence) technology to not only make the ride smoother for customers but also make track inspector’s jobs safer by equipping them with more advanced tools.”
Google Pixel smartphones were retrofitted onto R46 subway cars to capture subtle vibrations and sound patterns through built-in sensors equipped with a microphone. The data is sent in real time to cloud-based systems where AI and machine learning algorithms generate insights that indicate where preventive maintenance is needed.
NYCT track inspectors examine the locations highlighted by the system to confirm whether there is an issue.