The Car Speaks: Breakthrough Software Allows Your Car to Voice Its Complaints

New smartphone app can tell what is wrong with a car, just by listening to it.

Have you ever imagined stepping into a car and being able to tell everything wrong with it, just by taking a glance at your smartphone? From the right tire that needs air, to the air filter that needs to be replaced in a fortnight? With the new auto-diagnostic app by Josh Siegel, these cards may very well be on the table, as science fiction is gradually becoming science reality.

Josh Siegel: Developer of the auto-diagnostic app (courtesy: Slice of MIT)

Josh Siegel: Developer of the auto-diagnostic app (courtesy: Slice of MIT)

Although there are quite a number of automotive diagnostic tools such as OBD scanners, which read and interpret error codes from the vehicle’s systems, they usually require some sort of  special connection to the vehicle before they can be used, not to mention the need for technical know-how. But with this new app, no technical knowledge or special connection is required to detect whatever problem the vehicle is suffering from.

The app functions by listening to the car’s vibrations and sounds, through the phone’s microphone and accelerometers. It analyzes the sounds and vibrations by comparing them with the normal sounds and vibrations expected in faultless cars. It pinpoints components or system that are already faulty or will soon develop a faults and offers recommendations as to how and when such issues need to be fixed.

For most diagnostic scans, the smartphone simply needs to be in the car, but for certain diagnostics, mounting the smartphone to the dashboard holder will enhance accuracy. The app capable of detecting a wide range of issues, cutting across every system in the vehicle. For instance, it can tell when a tire is worn-out, deflated or over-inflated, or when the pores of the air filter is becoming too thin for the car to “breathe.”

The auto-diagnostic app could potentially save an average car owner, up to $125 a year and a truck driver up to $600 annually, not to mention the additional benefit of avoiding unaffordable breakdowns. This is principally because, the app provides preemptive warning concerning the vehicle’s conditions, what maintenance is required and when to carry out such maintenance, before such issues become real complications.

This app was designed in MIT by a team of researchers led by Joshua Siegel PhD, alongside professor Sanjay Sarma and the details of the research are being published in the November issue of Engineering Applications of Artificial Intelligence. A working prototype of this app is currently being developed and should be available for public testing in six months, after which a commercial version will be produced and released a year after.

For an in-depth look at the future of automotive technology, check out Driverless Cars – The Race to Level 5 Autonomous Vehicles.