The New York City Subway Is Using Google Pixels to Listen for Track Defects

Between September and January, six Google Pixel Smartphones have hits free rides in four New York City Subway cars. Specifically, they took a train, as it ping 32 miles in the middle of the north tip of Manhattan and the southern of Queens.

The phones were not stoway or stray and a very sharp eye passenger could say because they were inside the plastic enclosure and the vehicle was protected through brackets in the underside and interior. When people inside the cars are used to write their smartphones or to scroll on Instagram or explore roblox, subtile operators were using sensors of these phones – using acceleometers, magnetic and gyroscopes, and for attachments to the cars.

The phones were part of a short test in New York City Metropolitan transport authority And Google Cheap, most off-shelf technology agency’s track inspection can complement the work. (Google Public SectorThe section that started the work, did not charge the MTA for this initial exam.) Today, human inspectors have visited, who simultaneously walked 665 miles of the subway in the subway of New York City, leaving the eyes broken for problems like broken rail, bound signals and water loss. Three times annual rides by specialized, censor-bouquet “train geometry cars” also capture and upload more sophisticated data about the city’s rail infrastructure status.

Work with experimental technology in New York City Transit, which calls Google Trackinspect, suggests that audio, vibration and position data has been collected relatively cheaply and used for training Artificial intellect Forecast models, that inspection can complement the task. It can point people to suspicious rats, bans or skills, refer to what kind of equipment they need to repair them before they go there. Throughout the entire four -month project, the technology was later able to detect 92 percent of the defective position identified by human track inspectors, MTA.

Finally, the technology “can become such a way” we can reduce the amount of work done to identify those flaws and point the visitor to the right direction, so that they can spend time to fix and go straight to the work, “agency president Demetrius Crealo says, in the future, MTA will create a” modernization system and detects a “modernization system.

For the system of 1.7 million daily riders of the system, defects may be different from timely work or schooling and unexpected delay in the defects before the problem is.

“The target with this [project] Problems can be found before they have a major problem in terms of service, “Crisslow said. Cooperation with Google will now expand on a full pilot project, MTA, where Google will create a production version of technology and keep track inspectors on its own.

Inspectors’ gadgets

The Google Exam is part of the AI-enabled technology bumper crop that transit agencies have just begun to use their general inspections, with transit and rail assistant vice-president Brian Poston, with WSP. Although the New York “Harmonics” – is unique using audio and vibration using, others have installed small sensors or cameras on the tracks that accept automatic measurements and flag significance as soon as it grows. The technology enables the advanced and small batteries and processors not only by the advancement of machine learning.

Nevertheless, US regulators need regular humanitarian inspection and maintenance of the railway and the poston says he does not expect these rules to go away at any time. “You will always need that human interaction until the technology is precise and precise,” he said.