The Stanford Cart was a small wheeled robot developed at the Stanford Artificial Intelligence Laboratory, originally built to simulate a remote-controlled moon rover. In 1977, Hans Moravec rebuilt the Cart and equipped it with stereo vision: a television camera mounted on a rail on top of the Cart took pictures from several different angles, and a computer used them to gauge the distance to obstacles and steer around them. According to the Computer History Museum, “In 1979, Moravec’s Cart successfully crossed a chair-filled room without human intervention.”
The achievement was real but agonizingly slow. The Cart moved in short bursts, stopping to take pictures and grind through the vision computation, taking hours to cross a single room. That slowness was itself an important lesson: perceiving the world well enough to navigate it turned out to be the hard part of mobile robotics, far harder than the mechanics of moving.
The Stanford Cart is a direct ancestor of later autonomous vehicles, from Shakey’s reasoning-driven navigation to the DARPA Grand Challenge cars and modern robotaxis. Moravec went on to lead mobile-robot work at Carnegie Mellon and to write influential books on robotics and machine intelligence.