WALLY THE ROBOT HOW TO

However, once students were asked how they could make Irving undo its last move so as to get where it had been 'the time before last,' most had a rapid flash of understanding and knew how to complete the entire path reversal. But they didn't know how to use this for reversing Irving's path. They knew that Forward and Back are inverse operations, as are RightTurn and LeftTurn, and PenUp and PenDown.

The notion that there is an algorithm for accomplishing it was not obvious. Even bright students were initially perplexed. How could they control what they couldn't see? The task was compelling and, for all but the most sophisticated children, seemed non-trivial. Students were typically taken aback, because they could no longer see the turtle when it was in the adjoining room. The student had a complete record of the sequence of commands she had used to control Irving, since each command was printed on the teletype. After Irving was out of view, the challenge was to bring it back 'home,' to its starting point in the original room, only through the use of Logo commands, without peeking into the adjoining room, for example. This typically required a sequence of about twenty move and turn commands. The first part of the task was to move Irving from its initial room location to an adjoining room where it was no longer visible. However, we conceived a kind of limiting case, the 'path reversal' task. On the face of it this seemed impossible-intelligent behavior requires feedback from the environment. (If we had done that, we might have called Irving a beetle instead of a turtle!)ĭuring this phase of Irving's development, we wondered what kind of Al task, if any, was feasible for a robot without sensory feedback. Before we settled on bumpers as the appropriate device for touch sensors, we considered the use of antennas. However, Irving initially lacked touch sensors, so it was not able to detect obstacles. We were interested in seeing whether elementary and junior high school students could control Irving to do simple Al tasks such as avoiding obstacles while circumnavigating an area. It was capable of moving freely under Logo commands via a radio transceiver attached to a teletype terminal connected to a remote computer. In 1972, BBN engineer Paul Wexelblat designed and built the first wireless floor turtle, 'Irving.' Irving was a remote-controlled RF turtle about one foot in diameter. 'Screen turtles' were introduced at the MIT Logo Lab around 1972. Using Logo turtle commands, a student could command the turtle to move forward or back a specified distance, turn to the right or left a specified angle, sound its horn, use its pen to draw, and sense whether contact sensors on its antennas encountered an obstacle. As the turtle moved, it dragged the connecting wire along with it. In 1971, visiting consultant Mike Paterson introduced the Logo-controlled robot turtle, a 'floor turtle' physically connected to the computer via hardwire lines. Wallace Feurzeig, Principal Scientist, Learning Systems, BBNT, Cambridge, Massachusetts, USA

Perspective on Logo's Early Years, Legacy, and Toward a Culture of Creativity: A Personal

Wally the robot