Salto jumping robot adds precision hopping to its repertoire
We first saw UC Berkeley's remarkable jumping robot Salto in December 2016, but a new video released by the university's Biomimetic Millisystems Lab shows an equally remarkable new ability: precise hopping, pogo-style.
Salto stands for Saltatorial Locomotion on Terrain Obstacles, saltatorial being a Latin-derivded word meaning related to or adapted for leaping. This latest iteration of the robot is called Salto-1P and is specifically designed to hop between designated spots, as if jumping from stepping stone to stepping stone or playing hopscotch. It can hop on the ground, or between objects of increasing height.
The original Salto, inspired by the galago or bush baby, was capable of jumping higher and faster than any other robot thanks to its high power-to-weight ratio. A video at the time showed the robot chaining two jumps, allowing the robot to gain even more height than its initial 1-meter- (3.3-ft) leap from the ground allows.
What the 1P iteration adds, besides a greater number of consecutive jumps, is control, seemingly bringing Salto closer to its stated design aim of tackling difficult terrain in potential search and rescue situations. It uses what its creators call a "deadbeat foot placing hopping controller" for this extra precision.
The controller uses what's known in mathematics as a Taylor series approximation, which can be used to approximate answers to complex equations, including those governing forces acting on a point. The speed and direction of the hopping can be regulated by altering the way the robot lands. It also uses aerodynamic thrusters and an inertial tail to help control its altitude while in the air.
At the moment, Salto is still limited to jumping between pre-defined points. To truly search and explore, its jumping ability would probably need to be combined with vision and route-planning systems for it to be able to effectively explore unaided.
Work on 1P has been done by Duncan Haldane, Justin Yim and Ronald Fearing. The work is supported by the Army Research Office in the USA. You can see the new video below.