That 2012 paper from UC Berkeley on the advantages of giving mobile robots tails continues to inspire roboticists, nearly two years later. At IROS 2013, we checked out a new implementation of an actuated tail that makes for one seriously maneuverable robot car.
The robot is called Dima (a name derived from a Sotho word that means "flash of lightning"), and it was deliberately designed to be able to achieve high speeds with a high center of mass, a combination that works best (or only) when driving in straight lines. Turning at any speed that you might charitably call exciting leads to an immediate toppling over, but the addition of an actuated tail that can swing in the roll axis of the robot can effectively keep it stable:
This is a bit different from some of the other tail-assisted turns that we've seen robots doing (most notably this robot from UC Berkeley), because in this case, the tail is being used to counteract the torque that the robot generates while turning, rather than being used to generate torque to turn the robot. Also, the tail is being turned in a roll axis instead of a yaw axis, an idea that the researchers got after watching how a cheetah's tail moves when it makes high-speed turns.
After a bunch of experiments with both tail-less and tailed versions of Dima, results showed that the addition of the actuated tail allowed the robot to make stable turns at over twice the speed that it would be able to otherwise (7.5 m/s as opposed to 3.1 m/s).
So, the obvious question now is whether there are any rules in Formula One specifically prohibiting the use of active tails, and if not, when we'll start seeing them on race cars.
"Rapid Turning at High-Speed: Inspirations from the Cheetah's Tail," by Amir Patel and Martin Braae from the University of Cape Town, South Africa, was presented last week at IROS 2013 in Tokyo, Japan.
[ UCT Robotics ]