With public trust in vehicular autonomy remaining relatively shaky, and steady (but painstakingly slow) progress being made on weeding out the bugs and other flaws, safety is shaping up to be the x-factor in the future success of this innovative industry. Self-driving cars are riddled with drawbacks, including the operability of their sensors in bad weather conditions or at night, along with navigational errors like recognizing driving lanes and crossing bridges. Corporate entities in the automotive and other industries are working on solutions to mitigate or eliminate these faults among driverless cars, with many focusing on improving aspects like the vehicle’s software, sensors, or control systems.
Renault is among these companies, who have developed an autonomous vehicle control system capable of avoid obstacles at the equivalent speed of professional test drivers. Since successfully avoiding objects will be a necessary feature for any autonomous vehicle, this system was actually inspired by the abilities of professional test drivers, who were used as a benchmark for the level that the control system should be performing. While it’s unclear at what speeds the system can effectively operate, the video below shows a demonstration of an autonomous car equipped with this evasive technology going under 30 mph.
The avoidance capabilities mentioned were tested in Renault’s self-driving prototype called Callie, which navigated through an obstacle course that deployed inflatable barriers at the last second for the vehicle to surmount. As the video below shows, Callie eluded these popup obstacles with great ease, having to make some sharp but timely maneuvers. While there’s still a lot of work to be done, Renault seems confident that their successful tests with Callie are a great starting point.