In today’s cars more and more functions are implemented in software; most innovations in vehicle construction come from software. This confronts car manufacturers with ever shorter development cycles and frequent and unpredictable software problems. This, in turn, results in higher recall rates. In 2017, for example, 15 million vehicles were recalled due to software errors, costing billions of dollars. The predicted increase in the amount of software code in vehicles will reinforce this trend.
Aurora Labs offers a predictive maintenance solution for connected cars and autonomous vehicles. Its machine learning algorithms address all three stages of vehicle maintenance: The platform detects errors in software behavior and predicts downtimes, and it corrects errors in the ECU software. In addition, the clientless Over-the-Air (OTA) update solution from Aurora Labs offers fast ECU updates without downtime and without the double storage requirements otherwise associated with software updates over-the-air.
“The number of lines of code in vehicles is already around 150 million and is expected to increase further,” argues Aurora Labs CEO and co-founder Zohar Fox. “On average there are about 15 to 50 errors per thousand lines of code, 15 percent of which are overlooked by quality assurance. This highlights the need for solutions that can predict downtime before it leads to security problems.”
To date, Aurora Labs has three pilot projects underway with major automotive OEMs. Further projects are planned for the coming months. Founded in 2016 by Zohar Fox and Ori Lederman, the company has offices in Tel Aviv and Munich. Aurora Labs has developed a process called Line of Code Maintenance. Using machine learning, it can detect errors in the embedded software of the vehicles and repair them using OTA.
More information: https://www.auroralabs.com/