Between September and January, six Google Pixel Smartphones collected free rounds on four New York City Metro cars. Specifically, they took the Air A, as it extended 32 miles between the northern ends of Manhattan and the southern access of the queens.
Phones were not corridors or lines, and a very sharp passenger can say because they were inside the plastic packages and are secured through arches to the sides of cars and interiors. While people inside the cars used their smartphones to write emails or scroll on Instagram or Roblox exploration, subway operators used these phones – speedy speeds, magnetic standards, gropscope, and for those associated with car sensors, and additional external microphones – to listen in general.
The phones were part of a brief experience by New York City Urban Transport Authority and Google To whether cheap technology, mostly, can complement the proximity work on the agency. ((Google General SectorThe section that did the work, MTA did not receive this initial experiment.) Today, inspections are performed by the human inspectors, who together walk all the subway tracks in New York City, and the eyes of scales such as broken bars, broken signals, and water damage. Horse riding three times by specialized “train engineering cars”, as well as capturing and downloading more sophisticated data on the city’s railway infrastructure.
Transit’s work in New York City with experimental technology, which is called by Google Track Circuit, indicates that sound, vibration and location were collected relatively low and used for training artificial intelligence Prediction models, you can complete this inspection work. Humans can be directed to suspicious bells, explosions, or raying, which indicates the types of tools they will need to make repairs before arriving there. Throughout the four -month project, technology was able to determine 92 percent of the faults locations that the human path inspectors later identified.
Ultimately, the technology can become “a method that can reduce the amount of work that is done to determine those defects, and the points inspector in the right direction, so that they can spend some time in installing instead of defining and going there directly and doing the work.” In the future, MTA hopes to create a “updated” system that determines and regulates repairs to the paths of the path automatically.
For 3.7 million passengers per day for the system, picking defects before becoming problems may be the difference between work or school on time and confusion in an unexpected delay.
“The goal with this [project] “It is finding issues before it becomes a major service.” Crichlow says. MTA says, collaborating with Google will now expand to a full demo project, where Google will build a production version of technology and put it in the hands of the path inspectors themselves.
Inspectors tools
The Google experience is part of a technology that supports artificial intelligence that transport agencies began to use to supplement their typical inspections. Although New York is unique in the use of “AUDIO” and vibration – to determine problems, others have installed small sensors or cameras on the paths that take mechanical measurements and science contradictions when they appear. Technology is enabled not only through developments in machine learning, but also the cheapest and smallest batteries and processors.
However, the American organizers need a regular examination of the railway and human maintenance, and Buston says he does not expect these rules to disappear soon. “Until technology is specific and accurate, you will always need this human interaction,” he says.
