New Algorithm to Enhance Performance of Robots

New Algorithm to Enhance Performance of Robots


A speedy collision detection algorithm is likely to improve the performance robots that would now act more like a human assistant at home or in operating rooms. Electrical engineers of the University of California San Diego have come up with a program called ‘Fastron’. This algorithm, Fastron, runs 8 times faster than the prevailing algorithms of collision detection.

Applications of Fastron

The team that is led by a computer and electrical engineering professor, Michael Yip, is all set to present their new algorithm at the headquarters of Google at Mountain View, Calif. The algorithm will be revealed at the very first day of a 3-day long annual conference of Robot Learning. The team of engineers hopes that the algorithm “Fastron” will be widely used for robots in human environment settings wherein they can interact and work with people and moving objects smoothly. One of the applications that the engineers are looking into is surgery that is assisted by robots. Robot-assisted surgeries that make use of the Da Vinci Surgical System might use a robotic arm that would be tasked to autonomously do assistive tasks like irrigation, pulling tissue back or suction without intervening into the tasks of the surgeons.

The collision detection algorithms that are currently available is heavy with computations. With mobility of the objects, the computation becomes heavier and complicated. It is to lighten this burden of computation, Yip and his team came up with a minimalistic approach towards collision detection. His team has, so far, used the algorithm in simulations of computer on obstacles in simulation and robots. The team is further trying to enhance the accuracy and speed of Fastron.