Standard computing technology simply isn’t good enough to develop safe autonomy.
Since the introduction of Tesla’s Autopilot software, just about every OEM out there has scrambled to catch up. Now, if they’re not caught up, these OEMs have surpassed the EV brand’s software. Unfortunately, improving the software isn’t as easy as you might think. Computers don’t think like us, react like us, or make judgment calls like us. It’s just part of the way the world works.
However, a number of brands are trying to find ways around this, teaching machines to think more like a human driver than a binary, one-way, yes-or-no machine normally would. That includes Hyundai and by extension, Kia. The makers of the Hyundai Ioniq 5 and Kia EV6 have partnered with IonQ to improve their autonomous technology.
Hyundai/IonQ
Hyundai
Hyundai
IonQ, a quantum computing company, says technologies like image classification and 3D object detection are key to further developing autonomous driving technologies. The company will work with Hyundai, using machine learning to improve the functionality of these systems. Already, the pair of systems are working on classifying 43 different types of road signs using IonQ’s quantum processors. Road sign identification is critical for autonomous software and can help it “act” more like a person.
That also means the software needs to be able to identify road signs under a variety of conditions. We as humans know what a stop sign covered in snow looks like, but a computer only knows what a perfectly clear and visible one looks like. Machine learning and image classification are key to bridging that gap, says IonQ. “We expect to see quantum computers become an even more integral part in developing novel transportation solutions,” said Peter Chapman, President and CEO of IonQ.
The two companies aren’t just working on teaching machines what stop signs look like, however. They’re also working to help the software recognize other objects too, which is where the 3D imaging comes in. It’ll help autonomous software identify the difference between a tumbleweed and a soccer ball, for example. Object recognition tasks are being run on IonQ’s latest quantum computer, which should help enable more efficient processing and lower costs for both brands. The IonQ Aria is, at present, the most powerful quantum computer in the industry for standard application-oriented industry benchmarks.
Ideally, these results will soon trickle down into production vehicles. The public consciousness has inextricably linked EVs like the Ioniq5 with advanced technologies, and being able to deliver on that is critical to an automaker’s success in the electric vehicle segment.