Figure 1. Charles A. Rosen and the “Automaton.”
let me quote from one of these people, a junior in
high school at the time. He and a high-school friend
traveled to visit the Shakey project in 1971, unannounced, but were welcomed by the Shakey team.
I was inspired by the Shakey video from SRI. I actually went down and visited when I was a junior in high
school and they showed me the lab.
Shakey was pretty cool — vision, modeling, planning. It
decided to move things around so it could go up a ramp.
Paul — do you remember how we got this video?
The “Paul” is … Paul Allen; and the author of the
quote is Bill Gates.
Peter E. Hart
The proposal that launched the Shakey project was
submitted by the Artificial Intelligence Center of
Stanford Research Institute (now SRI International)
in January, 1965. SRI proposed to develop “
intelligence automata” for “reconnaissance applications.”
But the research motivation — and this was the inspiration of Charles A. Rosen, the driving force behind
the proposal — was to develop an experimental test
bed for integrating all the subfields of artificial intelligence as then understood. SRI wanted to integrate
in one system representation and reasoning, planning, machine learning, computer vision, natural
language understanding, even speech understanding, for the first time.
Readers interested in technical details of Shakey’s
development will find an excellent summary1 in an
SRI report. A 25-minute video, made by the Shakey
team at the time, is available. 2
The design of the “automaton,” as it was initially
called (perhaps out of a justifiable concern that
“robot” sounded like science fiction, which it was
before Shakey), was governed by two ground rules:
First, in order to keep it mechanically as simple as
possible, no arm was installed. And second, to avoid
issues of miniaturization, the design evolved as an
electronics rack on wheels with a sensor assembly
mounted on top.
The project team was well aware of Shakey’s limited mechanical and sensory capabilities, and designed
a correspondingly simple experimental environment
consisting of half a dozen rooms populated with
large, geometric blocks. The blocks were painted so
that edges were visible to the low-resolution TV camera, while still being sufficiently reflective for our
homemade laser rangefinder to work. We also used
dark baseboards, again for visibility, and exploited
them to update the position error that accumulated
in the dead reckoning process that relied on Shakey’s
Our first computer was an SDS 940, an early commercial time-shared mainframe (whose main memory was smaller than the L2 cache of most laptops). In
1970 we upgraded to a more powerful DEC PDP- 10.
Shakey talked to the PDP- 10 through a communications processor, and the system was one of the handful of nodes that constituted the birth of the
ARPANET. Around this time we embarked on a complete rewrite of much of Shakey’s software, while
making only minor upgrades to the robot hardware.
In the next section we describe this version 2 of