Researchers at Georgia Institute of Technology are working on ways to make robots smarter, and it appears they have come up with a most intriguing way of doing so.
The researchers have created arm sensors that are capable of reading a person's muscle movements. The sensors send information to the robot, which allows it to anticipate movements of a human and correct its own movement. This system is aimed at improving safety and efficiency in manufacturing plants.
Most robots are designed and programmed with sensors to keep them from accidentally injuring or killing any human who might come close to them. The sensors Georgia Institute of Technology is working on will enable robots to now predict human movements, which could enable both to work seamlessly alongside each other with a higher degree of safety.
"It turns into a constant tug of war between the person and the robot," explains Billy Gallagher, a recent Georgia Tech Ph.D. graduate in robotics who led the project. "Both react to each other's forces when working together. The problem is that a person's muscle stiffness is never constant, and a robot doesn't always know how to correctly react."
Shifting levers forward and backwards in a manufacturing plant cause robots to move accordingly. However, according to Georgia Tech, what happens when workers want to stop movement? They tend to stiffen when holding the lever in place, which is capable of confusing the robot. The robot has no way of telling whether that contraction is a command to change direction or just the hit force that results from the muscles contracting.
"The robot becomes confused. It doesn't know whether the force is purely another command that should be amplified or 'bounced' force due to muscle co-contraction," said Jun Ueda, Gallagher's advisor and a professor in the Woodruff School of Mechanical Engineering. "The robot reacts regardless."
By reading the person's muscle movement, the sensor can now determine the operator's actual status and movements, and send that information to the robot.
"Instead of having the robot react to a human, we give it more information," said Gallagher. "Modeling the operator in this way allows the robot to actively adjust to changes in the way the operator moves."
