Nissan has taken NASA technology built for the Mars Rover to help control fleets of autonomous vehicles. It’s called Seamless Autonomous Mobility (SAM), and Nissan believes it’s a solution for when vehicles don’t know how to handle the unexpected.
Though even basic self-driving car systems can handle the vast majority of situations — heading down a well-lit highway in nice weather — the biggest obstacle to fully autonomous driving is how to handle unusual or unpredictable or just plain weird situations.
SAM allows a “mobility manager” to examine vehicle images and sensor data when the car encounters something it can’t handle and decide on the appropriate course of action. Nissan uses the example of a police officer using hand signals to direct a vehicle around a car accident. If the cop wants the car to head the wrong way down a one-way street, the autonomous vehicle may not be confident in its assessment of the situation and request assistance.
The mobility manager would sit in a call center of sorts, able to take requests from entire fleets of vehicles. Once he assess the situation, he can “paint” a route for the vehicle to take. Once it’s past the obstruction, the car can resume autonomous operations. Best of all, the entire fleet can learns from the experiences and solutions will be shared between vehicles. Every time a solution is manually plotted, it reduces the likelihood that cars will need assistance in the future.
SAM is adapted from NASA’s Visual Environment for Remote Virtual Exploration (VERVE), which is used to supervise interplanetary robots like the Mars rovers. NASA scientists can use the system to calculate safe driving paths for on other plants, allowing the robot to execute the final maneuver.
Nissan thinks the system could be suitable for commercial fleets handling package delivery or taxi and ride-sharing services, and that it could help autonomous cars deploy much sooner than if we had to wait for autonomous systems to be able to handle every situation that might come along.