Good Vibrations: New Technology Analyzes Vehicle Sounds to Inform Consumers of Maintenance Needs
Wouldn’t it be great if you could know exactly what items in your vehicle needed fixed without visiting your local car shop? MIT researchers have been finessing a new software that would inform you of a car’s maintenance needs via a smartphone app.
The gist of how it works is that a smartphone uses its accelerometers and microphone to detect a vehicle’s vibrations and sounds. It then analyzes the results to compose a diagnostic report which you can view on your phone’s touchscreen.
For instance, one of the vehicle concerns the software can sense is when the car needs a new air filter. According to research scientist Joshua Siegel, PhD, a dirty air filter makes a distinct sound: “As it starts to get clogged, it makes a whistling noise as air is drawn in. Listening to it, you can’t differentiate it from the other engine noise, but your phone can.”
Right now, smartphones are sensitive enough to pick up signals that indicate basic maintenance items without being mounted to a vehicle. However, to detect certain maintenance issues, the smartphone would have to be connected to the dashboard via a mounted phone holder.
The new technology’s results currently have an accuracy over 90%. With additional tweaks, the team hopes to further improve the exactitude of the results.
With this innovative technology, consumers could more conveniently stay on top of car maintenance. The MIT research team hopes that it will enable people to detect vehicle issues right when they start, to help reduce the price they spend on repairs as well as increase their vehicle’s overall gas mileage.
When will this software hit the market? The MIT team is working on a prototype smartphone app that taps into the nascent technology; it plans to release it for field tests six months from now. The app will be publicly available about one year after the field testing phase, through a new startup called Data Driven, which Joshua Siegel, PhD, formed.
News Source: ScienceDaily