Companies like Tesla, GM, Uber, Ford, Google’s Waymo and many more spend lots of money developing autonomous drive systems. Once you factor in the teams of engineers, expensive sensors like Lidar, and high-end cameras, these self-driving vehicles become worth their weight in gold.
This massive need for capital could create financial woes and hurt the success of the project. But a UK-based startup says they can create an autonomous system for much less thanks to the help of AI, and released a video showing what it can do.
Wayve (not to be confused with Waymo) claims their autonomous drive system costs less than 10 percent to that of the competition’s. If accurate, that’s impressive. What’s more impressive is that Wayve’s system doesn’t rely on complicated and extensive rule-based software or an HD map of its environment captured by pricey cameras. Instead, its autonomous drive system uses a machine-learning AI platform, a navigation system, and simple cameras. Furthermore, everything operates on the modern-day equivalent of a laptop.
If that’s not enough, Wayve says the route its vehicle took in the video was the first time the system had seen it and without the use of a camera-generated HD map of its surroundings. A “world first” according to the company. But even if their claim is not completely accurate (not to say that it’s not), that is still a big accomplishment in the autonomous driving world.
“The traditional approach used by all our competitors relies on HD-maps, expensive sensor suites and hand-coded rules that tell the car how to drive. We have built a system that learns end-to-end with machine learning. It is the first in the world to drive on urban roads it has never been on before,” said Alex Kendall, co-founder and chief technology officer.
As neat as it is to see a car drive itself, you’ll notice in the video that the vehicle is usually only going about 8 or 9 mph and is on a mostly straight route with some simple turns. Examples of highway driving or complicated intersections are not shown in this video. However, at two points in the video, the autonomous car managed to avoid bicyclists that crossed its path.
If Wayve’s cost claims are accurate and their system can manage more complex driving situations, this could make autonomous systems more affordable and possibly get them to the market faster. If other companies haven’t introduced AI into their drive system yet, they might reconsider after watching this video.