AIMotive wants to scrap the idea that autonomous vehicles have to be built from the ground up, by developing technology that can be implemented into existing vehicles.
The company is adding regular cameras to cars and using its own developed artificial intelligence engine rather than equipping them with more expensive Lidar technology.
It builds a 3D landscape of the surrounding area using footage from the cameras streamed to a PC in the back of the car. The images are stitched together to create the model and referenced to a GPS map of the area. AI is then applied to the landscape, highlighting potential dangers, such as objects, pedestrians and anything else that needs to be avoided.
The company thinks this will reduce costs for developing automated cars from between $70,000 and $100,000 down to the more manageable $6,000.
The company said it wanted to use less expensive and more humanlike technology because it wanted to fit autonomous technologies in with how humans think when they’re driving rather than making it even more robot-like.
“The whole traffic system is based on the visual system,” explained founder and CEO Laszlo Kishonti. “Drivers don’t have bat ears and sonars, you just look around and drive. The only way to do this is with AI.”
The company explained one challenge with the technology is that it can only react to what it can see, rather than remote detection of objects as is the case with Lidar, which uses radar to detect objects. However, this could also play to its advantage, because many cars using Lidar need a map of the area to reference, whereas AIMotive’s solution builds the map as it goes.
AIMotive’s tech could be implemented in both cars and trucks, which, the company thinks will account for the majority of automated driving experiences in future.