Today, self-driving car technology isn’t cheap--and rest assured that even when prices do go down, automakers will still charge a premium for the feature. But researchers working on the University of Oxford’s RobotCar U.K. project think they can make the technology so inexpensive (in the range of $150) that even a marked-up self-driving car will still be affordable.
RobotCar hasn’t created it’s own car; the initiative is working with a Nissan Leaf EV that contains lasers (under the front and rear bumper) working in concert with stereo cameras to create a 3-D image of the route being driven. A main computer controls the whole system, and an iPad mounted within reach of the driver communicates with a different computer (the Low Level Controller).
Here’s a video of the lasers sussing out the 3-D structure of a vehicle’s environment:
And here’s one of the stereo cameras interpreting live images in the context of memory of previous routes driven.
At the moment, RobotCar researchers are aiming to get the car driving short distances on familiar routes. PhysOrg explains:
During the initial period the RobotCar builds a 3-D map of routes taken by the car, which is being driven normally. Once it has built up a memory bank of familiar routes it begins to offer to take over and the driver can accept by tapping on the iPad screen. If there is any conflict between the three on-board computers autonomous driving is not offered. If there are any problems during autonomous driving the iPad prompts the driver to resume control. If they do not, the car slows to a stop.
At the same time, the laser located under the front fender scans ahead for obstacles. The car stops if anything is detected, and continues when the obstacle has passed.
The team estimates that its system could be ready for commercial production in just 15 years. The technology costs about $7,750 today, but as mentioned previously, researchers expect it to drop to $150 in the future. As for Google? Its sensors reportedly cost at least $250,000.