A new app from the U.K. uses the sensors in rail passengers’ smartphones to detect and map every bump, shimmy, and jiggle as their train rolls over defects in the track. The researchers behind the app hope that it could be used to make a virtual map of the country’s rail system, which could then guide repairs to improve the safety and comfort of the network.
The research comes from the University of Birmingham, and relies on the bevy of sensors in the modern smartphone: Accelerometers, gyroscopes, and GPS work in concert to detect and record sways and bumps, and place them accurately on a map. The system uses a neural network to combine the data from all the phones being used, to create an overall map.
The experimental app’s schtick is that passengers could use it to report comfort levels of their ride, but the data would be invaluable as a way to monitor the health of the railway infrastructure. Maintenance could be planned to hit the worst spots, or the busiest, improving safety as well as preventive further damage.
It turns out that today’s smartphone is more than up to the task. The researchers ran tests alongside reference accelerometers, and found that the phone sensors were just as good, and in the near future, they will probably be even better than the purpose-built test equipment.
This isn’t the first time that regular phones have been used this way. Back in 2015, an app called Street Bump launched to detect potholes in Boston, using the accelerometers and GPS to automatically report when a car hit a big bump. And any user of Waze is familiar with how powerful the network effect can be: Waze is a car navigation app that tracks the movement of everyone using it. This allows it to identify traffic snarls and route its users around them.
But Waze has a clear benefit to users. They let the service track them in return for drastically shorter journey times. Why would anyone allow their phone to track the bumps and shakes of their train ride? The researchers have already thought about that, and propose that the app be tied into the trains’ onboard Wi-Fi. You’d access free internet through the app, and it would provide data in return. Also, the privacy issues are lesser on a train than in a car: Unless something goes very wrong, you’ll be following the same known route as every other passenger on board.
This kid of big data is invaluable. Waze already Strava shares billions of data points that show cities how cyclists move through them. There will always be privacy and security tradeoffs, but the benefits can be huge.