answers they were seeking, and we can learn from the
blind alleys they spent time in, and the insights that
led them to the right paths.
Artificial intelligence marks its birth at the 1956
Dartmouth Conference. There have been many
important milestones along the way. The important
milestone we will celebrate today is the Shakey project, which created a physical robot that could perceive its environment and the objects within it.
Shakey could make a plan to achieve a goal state. And
it could carry out that plan with physical actions in
the continuous world. The Shakey project laid a
foundation for decades of subsequent research. We
are here to celebrate and understand that project.
The centerpiece of the Shakey celebration was a
panel presentation at AAAI- 15, designed to give the
audience an understanding and appreciation of the
process of the research in the Shakey project, and of
the long-term impact of that work on the larger field
of AI. The goal was to have three speakers address ( 1)
the state of the art in AI before the Shakey project (Ed
Feigenbaum); ( 2) the progress of the Shakey project
itself (Peter Hart); and ( 3) the impact of the Shakey
project on the future of AI (Nils Nilsson).
Celebrating Shakey and Its Builders
Edward A. Feigenbaum
The history of science is a source of knowledge of the
complex search for solutions to difficult problems.
Not only is this history endlessly intriguing and awe-inspiring; but also it should be of particular interest
to AI scientists because this kind of complex problem
solving and discovery is at the heart of many of our
theories of mental activity.
Life is lived in the moment. Everything else is
memory and stories. The word history itself contains
the word story. This talk is constructed as several stories of the Shakey project situated in its time, and
among other landmark AI projects.
I have been lucky enough to have lived and
worked through the entire 60 years of AI, from early
1956, months before the famous “founding” Dartmouth Conference, to today’s AAAI-2015. My stories
are drawn from those 60 years of memories, helped,
but only a little, by the best memory assistant ever,
My role today is to set the historical context in
which the Shakey project was born, lived a remarkable but short life, and was terminated. Shakey
research set the stage for decades of important experimental work in AI and robotics, and in other AI
applications that will be mentioned later by Nils Nilsson.
I phoned several well-known robotics scientists to
ask about the grandchildren of Shakey. All of them
said the robots they developed were grandchildren of
As shown in an original Shakey video, we remem-
ber Shakey as slowly and laboriously computing
models of its environment; planning; moving and
navigating its way around obstacles toward a goal on
the far side of one large room.
Fast forward to some recent news about grandchil-
dren of Shakey, from Manuela Veloso at Carnege Mel-
lon University (CMU):
I am very pleased to tell you that today, on November
18, 2014, the CoBot robots ( 3 of them) have jointly
autonomously navigated for 1,000 km in our multi-
floor SCS buildings at Carnegie Mellon University!
A great-grandchild of Shakey, Stanford’s self-driving car Stanley, the car that drove itself across the
Mojave Desert, is in the Smithsonian National Air
and Space Museum in Washington, DC. Other cars
like Stanley, built at Google, have driven more than
700,000 miles, navigating the San Francisco Bay Area
and other roads, according to the San Jose Mercury
News of November 12, 2014. (Consider this: Shakey’s
traversal, integrated over the whole life of the experiment, probably never made it to one kilometer).
Shakey’s grandchildren on Mars are still having a
productive long life — 11 years into a planned 90-day
visit, semiautonomously assisting planetary scientists.
I would now like to tell you personal stories that
together made the importance of Shakey research
vivid to me.
In 1993, a major Japanese corporation asked me to
do an evaluation of the quality of a robotics project
that its research lab been working on for several
years. After signing a nondisclosure agreement, I was
shown a robot that was “humanoid,” but very big
(scary, actually). Tethered to a power source, its
motion was fluid, a marvel of modern electromechanical engineering.
Though heavy, it could walk reliably without
falling, and it could even climb a flight of stairs. But
this creature had no Mind. It did no symbolic processing, no problem solving. It did not have goal-directed behavior.
There was more than enough space inside for a PC-sized computer and there was plenty of power. What
this project lacked were scientists and engineers
trained in AI, or even trained in software systems.
There were no young Nils Nilssons, no young Peter
Harts, no young Bert Raphaels or Richard Fikes —
and of course no visionary like Charles Rosen to integrate AI with electromechanical engineering.
And this was 1993, twenty years after the end of the
Shakey project! It can be perilous to ignore scientific
The Computer History Museum in Mountain View,
California, is the world’s premier museum for the history of computers and information technology and is