Driving 15mph over a curb 30 times a day can be tough on the drivers who perform vehicle durability tests, so Ford Motor Co. developed a better idea: Let the vehicles drive themselves.
Since late last year, the automaker has been using robots to drive trucks at its Michigan Proving Grounds, especially in jarring or physically demanding tests. "We're using the robots on the really tough roads," Dave Payne, manager of vehicle development operations for Ford, told us. "The drivers take a terrible beating on those roads, so we are trying to take them out of the equation and let the robots take the abuse."
Click the photo below for a look inside Ford's self-driving test vehicles.
Ford's robotic driving system includes a ring gear attached to the steering wheel, a shifter actuator
(in red), and an accelerator actuator (next to the driver's seat).
Ford is believed to be the first automaker to acknowledge it is employing robot drivers. Its engineers, who co-developed the robots with Autonomous Solutions Inc., use an electronic controller about half the size of a desktop tower PC. Sitting in the truck's back seat, the controller communicates with a GPS system that knows the truck's location on the 4,000-acre test facility. If the vehicle strays from a preprogrammed course stored in memory, the controller can send messages to a motor-based actuator that turns a ring gear attached to the steering wheel, enabling the vehicle to steer back to its desired location. If the preprogrammed test calls for the truck to accelerate, brake, or move its shifter, the system's linear actuators can do that, too.
"It basically operates like a human would, pushing on the pedals or turning the steering wheel," Dru Brown, marketing manager for Autonomous Solutions, told us.
Though the robot driver doesn't possess the autonomous capabilities of the well-known Google car, it can detect obstacles such as inattentive humans. An onboard optical sensor can recognize the presence of humans in front of or behind the truck and actuate the brakes. "Just in case someone goes in the area where I told them not to go, the camera will always pick them up and shut the vehicle down," Payne said.
Using onboard data acquisition capabilities, the system can also gather data, which has shown the system is more repeatable than humans, since it performs tests identically every time. "The more repeatable the data, the better the tool. If you're changing components, and you want to know if your change had an impact, you want to take out as many variables as you can."
According to Payne, Ford's technology is far less costly than Google's. "I don't know what their price tag is, but it must be huge. We're in the $100,000 price range. If it was more, it wouldn't make business sense."
Ford is employing the technology in a variety of tests on different driving surfaces, including broken concrete, cobblestone, metal grates, rough gravel, mud pits, and oversized speed bumps. The automaker has already used in it on the Transit, Super Duty, F-150, Navigator, Expedition, Fusion, and Fiesta vehicles.
Autonomous Solutions, which previously employed its technologies on agricultural and mining equipment, expects to see more applications like Ford's in the next few years. "It's dramatically improving the safety on some of Ford's test courses," Brown said. "We absolutely anticipate that other automakers will pick this up."