Robot Discovers Itself, Adapts to Injury

Nov 16, 2006
Robot Discovers Itself and Adapts to Injury
Graduate student Viktor Zykov, former student Josh Bongard, now a professor at the University of Vermont, and Hod Lipson, Cornell assistant professor of mechanical and aerospace engineering, watch as a starfish-like robot pulls itself forward, using a gait it developed for itself. the robot's ability to figure out how it is put together, and from that to learn to walk, enables it to adapt and find a new gait when it is damaged. Credit: Lindsay France/Cornell University

Nothing can possibly go wrong ... go wrong ... go wrong ... The truth behind the old joke is that most robots are programmed with a fairly rigid "model" of what they and the world around them are like. If a robot is damaged or its environment changes unexpectedly, it can't adapt.

So Cornell researchers have built a robot that works out its own model of itself and can revise the model to adapt to injury. First, it teaches itself to walk. Then, when damaged, it teaches itself to limp.

Although the test robot is a simple four-legged device, the researchers say the underlying algorithm could be used to build more complex robots that can deal with uncertain situations, like space exploration, and may help in understanding human and animal behavior.

The research, reported in the latest issue (Nov. 17) of the journal Science, is by Josh Bongard, a former Cornell postdoctoral researcher now on the faculty at the University of Vermont, Cornell graduate student Viktor Zykov and Hod Lipson, Cornell assistant professor of mechanical and aerospace engineering.

Instead of giving the robot a rigid set of instructions, the researchers let it discover its own nature and work out how to control itself, a process that seems to resemble the way human and animal babies discover and manipulate their bodies. The ability to build this "self-model" is what makes it able to adapt to injury.

"Most robots have a fixed model laboriously designed by human engineers," Lipson explained. "We showed, for the first time, how the model can emerge within the robot. It makes robots adaptive at a new level, because they can be given a task without requiring a model. It opens the door to a new level of machine cognition and sheds light on the age-old question of machine consciousness, which is all about internal models."

The robot, which looks like a four-armed starfish, starts out knowing only what its parts are, not how they are arranged or how to use them to fulfill its prime directive to move forward. To find out, it applies what amounts to the scientific method: theory followed by experiment followed by refined theory.

It begins by building a series of computer models of how its parts might be arranged, at first just putting them together in random arrangements. Then it develops commands it might send to its motors to test the models. A key step, the researchers said, is that it selects the commands most likely to produce different results depending on which model is correct. It executes the commands and revises its models based on the results. It repeats this cycle 15 times, then attempts to move forward.

"The machine does not have a single model of itself -- it has many, simultaneous, competing, different, candidate models. The models compete over which can best explain the past experiences of the robot," Lipson said.

The result is usually an ungainly but functional gait; the most effective so far is a sort of inchworm motion in which the robot alternately moves its legs and body forward.

Once the robot reaches that point, the experimenters remove part of one leg. When the robot can't move forward, it again builds and tests 16 simulations to develop a new gait.

The researchers limited the robot to 16 test cycles with space exploration in mind. "You don't want a robot on Mars thrashing around in the sand too much and possibly causing more damage," Bongard explained.

The underlying algorithm, the researchers said, could be applied to much more complex machines and also could allow robots to adapt to changes in environment and repair themselves by replacing parts. The work also could have other applications in computing and could lead to better understanding of animal cognition. In a way, Bongard said, the robot is "conscious" on a primitive level, because it thinks to itself, "What would happen if I do this?"

"Whether humans or animals are conscious in a similar way -- do we also think in terms of a self-image, and rehearse actions in our head before trying them out -- is still an open question," he said.

Source: Cornell University

Explore further: An android opera: Japan's Shibuya plots new era of robot music

add to favorites email to friend print save as pdf

Related Stories

Metal made like plastic may have big impact

Oct 09, 2014

Open a door and watch what happens—the hinge allows it to open and close, but doesn't permanently bend. This simple concept of mechanical motion is vital for making all kinds of movable structures, including ...

Fingerprints of life on Mars

Oct 08, 2014

NASA's Astrobiology Institute (NAI) announced that the SETI Institute has been selected as a new member of the NAI for a 5-year research program, "Changing Planetary Environments and the Fingerprints of Life." ...

Robotic solutions inspired by plants

Oct 06, 2014

EU researchers are demonstrating revolutionary robotic techniques inspired by plants, featuring a 3D-printed 'trunk', 'leaves' that sense the environment and 'roots' that grow and change direction.

Pressing the accelerator on quantum robotics

Oct 06, 2014

Quantum computing will allow for the creation of powerful computers, but also much smarter and more creative robots than conventional ones. This was the conclusion reached by researchers from Spain and Austria, ...

Recommended for you

Robots recognize humans in disaster environments

1 hour ago

Through a computational algorithm, a team of researchers from the University of Guadalajara (UDG) in Mexico, developed a neural network that allows a small robot to detect different patterns, such as images, ...

Japan toymaker unveils tiny talking, singing humanoid

Oct 15, 2014

Japanese toymaker Tomy on Wednesday unveiled a multi-talented humanoid robot, named "Robi jr.," which can converse using some 1,000 phrases and belt out about 50 songs, as well as move its limbs and head.

Can we teach robots right from wrong?

Oct 14, 2014

From performing surgery and flying planes to babysitting kids and driving cars, today's robots can do it all. With chatbots such as Eugene Goostman recently being hailed as "passing" the Turing test, it appears robots are ...

User comments : 1

Adjust slider to filter visible comments by rank

Display comments: newest first

HackerMike
not rated yet Aug 04, 2009
This is a concept that I've thought about and wanted to develop: a neural net that learned to maximize velocity. It would need to adapt to whatever appendages it had available to it. The problem with extending this idea is it's easy to define walking = maximizing velocity, but what about more elaborate actions? What rule do you apply for it to, say, avoid bullets or move when a gun is pointed at it? Higher levels of programming will be needed. Ideally, heirarchical neural nets must be developed:
gun detection
human detection
human-using-gun detection
walking (needed to get out of the way if all prior conditions exist -- as developed)

So, these various nets must be specifically taught and the rule to walk when all condition are met (so it can get out of the way). Their adaptive walking robot is great, but doesn't come close to building multiple, disparate nets for the robot to truly evolve its thinking.

Herein lies the problem!