MTA and Google Team Up: Subway Cars Get Smart with Pixel Phones for Track Inspections

New York’s MTA Partners with Google for Subway Track Safety

The Metropolitan Transportation Authority (MTA) in New York, in collaboration with Google, has embarked on an innovative project to improve track safety and maintenance. The initiative involves attaching Google Pixel smartphones to subway cars to collect data on track conditions.

Context and Goals

• Volume of Commuters: The New York City subway serves over 3 million people daily, making human-powered inspections challenging and less efficient.
• Automation Efforts: The MTA aims to automate the process of identifying and addressing track defects to enhance safety and maintenance.
• Project Name: Operation TrackInspect

Methodology

• Devices Used: Off-the-shelf Google Pixel smartphones were used to collect audio, movement, and geographic data as trains moved through the subway system.
• Data Collection: The project gathered 335 million sensor readings and 1,200 hours of audio recordings.
• AI Training: Data was fed into AI models to train them to identify track defects and other issues.

Results

• Defect Detection: 92% of recorded defects were confirmed by human inspectors, with an 80% success rate in analyzing vibration recordings.
• Model Development: Around 200 individual AI models were developed as part of the project.

Future Outlook

• Implementation: The MTA hopes to expand this technology using specialized hardware and more advanced AI models.
• Cost Efficiency: The project demonstrates that existing consumer-grade devices can be effectively used for critical infrastructure monitoring at a low cost.