14 DAY TRIAL // A team of investigators at the Georgia Institute of Technology (Georgia Tech) announces the development of a new method for allowing robots and their human operators to work together. The approach eliminates one of the main hurdles facing this type of cooperation, namely the fact that robots cannot properly react to human actions.
This problem was tackled by developing a wearable sensor that a robot's human operator carries on their forearm. The instrument reads the amount of muscle contraction that the person is producing, and feeds these data into the robot's processing unit.
According to Georgia Tech experts, this type of sensors will allow machines to predict, or anticipate, human movements around them, therefore becoming capable of correcting its own actions and responses. This would be a great asset to have in a factory where large robots and humans work in close proximity to each other.
In most industrial settings, the areas where large machines are at work are not occupied by humans, due to widespread safety hazards. However, in some factories, there is simply no way to avoid interactions between man and machine. Such is the case, for example, with automobile manufacturing plants.
In these facilities, human operators control a robot that lifts car doors or other heavy parts, for easier installation. However, a tug of war can easily occur between the two, since robots are not always sure about what is required of them. The control levers used to issue commands such as “move forward” or “move backwards” are not likely to clear up this mystery for robots.
The new sensors, developed by Georgia Tech robotics graduate student Billy Gallagher, will eliminate this silent conflict by relaying human muscle data to robots. The machines will then be able to figure out when their operators want them to stop, as opposed to when they should just keep going.
“It turns into a constant tug of war between the person and the robot. 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,” Gallagher says.
“You don’t want instability when a robot is carrying a heavy door,” comments Gallagher's advisor, Georgia Tech Woodruff School of Mechanical Engineering professor Jun Ueda.
The new sensors were developed with support from a US National Science Foundation (NSF) National Robotics Initiative grant. The team says that their new sensor could benefit human-robot interactions in the automobile, aerospace and military industries.
“Instead of having the robot react to a human, we give it more information. Modeling the operator in this way allows the robot to actively adjust to changes in the way the operator moves,” Gallagher says.
“Future robots must be able to understand people better. By making robots smarter, we can make them safer and more efficient,” Ueda concludes.