The MTA has been struggling with the maintenance of its subway tracks, with issues such as strange noises and flammable debris being common problems. To address this, the MTA has partnered with Google to test an AI-powered track inspection system using Google Pixel phones. The system, called TrackInspect, uses machine learning algorithms to detect defects in the subway tracks. The phones were strapped to subway cars and collected data on the tracks, which was then sent to cloud-based systems for analysis. The system was able to detect 92% of the defect locations that inspectors found. The MTA hopes that this technology will help to improve the safety and efficiency of the subway system.