A Cambridge-based start-up believes machine learning software is the key to autonomous vehicles and Wayve is developing machine learning algorithms for autonomous vehicles. Wayve, which includes the chief scientist at Uber amongst its investors, believes the industry has been doing too much hand-engineering and too little machine learning. The firm is hiring for positions in its Cambridge-based headquarters. "The missing piece of the self-driving puzzle is intelligent algorithms, not more sensors, rules and maps. Humans have a fascinating ability to perform complex tasks in the real world, because our brains allow us to learn quickly and transfer knowledge across our many experiences.
Engineers have taught an AI the basics of driving in '15 to 20 minutes' – a process that can take some humans dozens of hours behind the wheel. Wayve, which was founded by researchers from Cambridge University's engineering department, used a technique known as'reinforcement learning' to achieve the feat. This teaches the algorithm using trial and error, with correct decisions rewarded with uninterrupted driving, and mistakes being corrected by a safety driver in the car. As the test progressed, the algorithm behind the wheel learnt not to replicate any mistakes that had been corrected by the human safety driver in the car. According to the Wayve team, the AI learnt to drive and corner while staying inside its own lane within '15 to 20 minutes' after it first took to the roads.
Why weigh down a self-driving car with a lot of sensors, HD maps, and equipment when you don't have to? It claims it only needs a camera, GPS tracker, and a powerful computer to be able to drive anywhere autonomously. But experts who specialize in sensing technologies like light-based LiDAR and radar say the idea mostly comes across as preposterous -- or the very least, short-sighted. Most self-driving cars decide how to drive down a street as it happens -- picking up information about debris in the way, pedestrians on the sidewalk, the sun starting to set in the distance. Wayve doesn't try to interpret that much data since it can't really pick up much from its cameras.
As Elon Musk made clear Monday, the technology most of his competitors in the self-driving car space use to help vehicles detect what's around them is lame. And his option is way better. "LiDAR is a fool's errand," he quipped about the laser-emitting tool that, in the simplest terms, acts as eyes for autonomous cars. "Anyone who is relying on LiDAR is doomed." That's pretty much most of the businesses testing self-driving cars, including Waymo and Uber who went to court over LiDAR technology last year.