Stanford Research Institute Builds Shakey the Robot

Robots that can reason might be the stuff of sci-fi nightmares (and current realities). But in 1966, at their infancy, they seemed much more benign. Shakey the Robot, perhaps the first mobile robot to reason about its actions, looked like a photocopier on wheels and it moved blocks. If anything, it's a bit cute.

Led by neural-network researcher Charles Rosen, a team at the Stanford Research Institute (SRI) spent six years building Shakey from 1966 to 1972.

Rosen's team consisted of Nils Nilsson, Bertram Raphael and Peter Hart (who later led it), with Richard Duda, Richard Fikes and Helen Chan Wolf. Guided by a video camera and a remote mainframe, Shakey took instructions in ordinary English and pushed blocks between rooms.

To plan its routes, Hart, Nilsson and Raphael devised the A* search algorithm, published in 1968 and still used to this day to steer characters in video games.