Smartphones could be used to diagnose problems with a car just by listening to the vehicle’s sounds.
Using the phone’s existing microphone and accelerometer, an app could pick up noise and vibrations to alert a user to issues like a low tyre, an air filter in need of replacing and even the need for new spark plugs.
Engineers behind the futuristic idea say the tech could be ready within two years – delivering diagnostic information about any car a user is travelling in within minutes and saving drivers about £95 a year.
All the information would be picked up via analysis of the car’s sounds and vibrations, foregoing the need to sync a particular car to the app.
Research scientist Joshua Siegel, from MIT, is one of the co-authors of the latest paper on the tech published in November’s Engineering Applications of Artificial Intelligence. He has also set up Data Driven to commercialise the idea.
He explained that existing kit contained in smartphones makes the application possible.
“The sensitivity is so high, you can do a good job (of detecting the relevant signals) without needing any special connection.”
The team has recently been focusing on detecting when an air filter needs to be changed. But what does this sound like?
“We’re listening to the car’s breathing, and listening for when it starts to snore,” Siegel explains.
“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.”
In order to test a plethora of cars, the team ended up renting different makes and models to which they added a problem and then fixed it.
Sigel added: “I’ve rented cars and given them new air filters, balanced their tyres, and done an oil change.”
For air filters, diagnosis is a listening task by the smartphone, but in other cases it’s a multistep process.
To tell if a car’s tyres are getting bald or overinflated, the phone would use its inbuilt GPS to establish the car’s speed. Vibration data can be used to determine how fast the wheels are turning.
That can be used to calculate the wheel’s diameter which can be compared with the diameter that would be expected if the tyre were new and properly inflated.
Through machine learning, algorithms designed to detect wheel balance problems did a better job at detecting imbalances than expert drivers from a major car company, Siegel says.