Nobody enjoys a surprise car breakdown. Wouldn’t it be great if your car could send notifications about a potential problem before it malfunctions? Thanks to Israeli startup 3DSignals, vehicles wired with artificial intelligence (AI) might be able to send early warning signs to its drivers.
Utilizing the technique known as deep learning, AI studies the specific noise patterns that correspond to machine errors. This creates a predictive model that solves mechanical problems before any serious consequences arise. 3DSignals hopes to use this technology in automotive factories and within cars themselves. Recently, the company expressed interest in the autonomous car field, hoping to detect vehicle-related problems in future driverless taxis.
According to Yair Lavi, co-founder and head of algorithms for 3DSignals, “If you’re a passenger in a driverless taxi, you only care about getting to your destination and you’re not reporting maintenance problems. So actually having the 3DSignals solution in autonomous taxis is very interesting to the owners of taxi fleets.”
3DSignals uses ultrasonic microphones, detecting sounds up to 100 kilohertz. To put this in perspective, humans can only hear between 20 hertz to 20 kilohertz. The microphones connect to a computing device for processing, and the device eventually uploads all the recorded data to an online network. This is where the deep learning algorithms finally showcase their talents.
Few companies specialize in developing AI precisely listening for acoustic signals. With proper training, the company’s software can detect and predict certain mechanical problems with 98 percent accuracy.
“It’s important for us to be specialists in general acoustic deep learning, because the research literature does not cover it,” Lavi says.
To learn more about 3DSignals, click on the video below.