AutoX Self-Driving Car Uses $50 Off-the-Shelf Cameras
AutoX is a self-driving car startup that has flown under the radar. Until now. AutoX released the first video of its self-driving Lincoln MKZ that uses $50 off-the-shelf cameras from Best Buy to navigate the streets of California.
San Jose, Calif.-based AutoX is a self-driving car startup that has flown under the radar. Until now. AutoX released the first video of its self-driving Lincoln MKZ that uses only a handful of $50 off-the-shelf cameras from Best Buy to navigate.
Most companies developing self-driving cars rely on Lidar to create 3D maps of its environment, including market leader Waymo. Tesla has built its self-driving technology around a camera-based system, but it also uses a forward-facing radar, 360-degree ultrasonic coverage, and high-precision GPS.
AutoX is different in that it uses only cameras - there’s no radar, sonar or GPS. And after building the prototype self-driving car in house over six months, the result is quite impressive. The video demo atop this page shows AutoX’s self-driving car flawlessly complete four different drives, including trips at night, in rain, and on a winding road. Tesla Autopilot still struggles with winding roads).
One major advantage camera-based systems have over Lidar is a much lower cost. Waymo, for example, uses three Lidar systems on its hybrid minivans. But until a few years ago, a single Lidar unit cost $75,000. Waymo originally used Lidar built by Velodyne, but Waymo brought everything in house in December 2015 to ensure it can scale. A disadvantage of a camera-based self-driving car is that it collects less data to make decisions.
How AutoX Got Started
AutoX was founded by CEO Jianxiong Xiao, who is also known as Professor X. He left his professorship at Princeton and his post as the founding director of the school’s Computer Vision and Robotics Labs in the spring of 2016 to create what is now known as AutoX.
While studying at MIT, Xiao developed a system that taught self-driving systems to observe images from Google Street View and identify important features.
Xiao says future iterations of AutoX’s technology will support ultrasonic sensors and LIDAR to enhance driver safety. But for now AutoX is focused on building the cheapest, most accessible system possible for self-driving cars.
AutoX has about 20 employees, most of them computer vision experts. Xiao says AutoX isn’t interested in manufacturing cars and going head-to-head with the likes of Tesla. AutoX will look to partner with automakers to build the brains for self-driving cars.
Xiao also didn’t provide a timeline for the commercialization of AutoX’s technology, which seems to be a no-no in self-driving car circles. But AutoX better hurry. There are now 26 companies testing self-driving cars on California’s public roads.