ngential line of thinking. The problem with AD is the environments in which cars have to operate contains an almost infinite number of variables - even if the basic pattern of driving is fairly straightforward. It's the variables that make it hugely complex for machines to drive safely. If Wayve can make it work, however, it would dramatically reduce the costs and development time for making autonomous driving a reality.\" - Peter Richardson (Source: Slash Gear)","isAccessibleForFree":"False","hasPart":{"@type":"WebPageElement","isAccessibleForFree":"False","cssSelector":".paywall"}}1c:T4f0,

"Wayve claims to be approaching the autonomous driving problem from a different angle than other autonomous driving (AD) companies. Wayve wants to do away with expensive sensor arrays, 3D maps, and massive computing power. It is instead relying on relatively simple computer vision and deep learning models. I can't help feeling that Waymo et al., must also have considered this as an approach, even if only as a tangential line of thinking. The problem with AD is the environments in which cars have to operate contains an almost infinite number of variables - even if the basic pattern of driving is fairly straightforward. It's the variables that make it hugely complex for machines to drive safely. If Wayve can make it work, however, it would dramatically reduce the costs and development time for making autonomous driving a reality." - Peter Richardson

Subscription required to access this content