Adding meat to the bones: how LiDAR is at the heart of AV perception

Any road operator, human or not, must understand the surrounding environment. Though significant work has gone into building complex sensors and cameras for autonomous vehicles (AVs), ‘seeing’ is very different to ‘understanding’. Whereas a human can both see and provide context to almost anything in view—for instance, distinguishing between an actual vehicle and the image of a vehicle—training Artificial Intelligence (AI) to do the same is tricky. AI needs context and in the realm of the AV that requires a detailed three-dimensional foundation.