As robots get smarter and more capable and make their way from manufacturing assembly lines to a much wider variety of applications, we will be interacting with them in more and more situations. Currently, robots tend to move with jerky, stop/start motions, which can make it difficult for humans, who are accustomed to the fluid and dynamic movements of other humans, to easily recognize what the robots are doing. In an attempt to create robots that can better interact with humans, researchers at the Georgia Institute of Technology are getting robots to move in a much more human-like way.
Largely thanks to a generalization of research findings by researcher Albert Mehrabian, who conducted studies in the 1970's looking at the relative importance of verbal and non-verbal messages, it is a commonly held myth – often spouted at sales seminars – that up to 93 percent of communication effectiveness is determined by nonverbal cues.
GET 20% OFF A NEW ATLAS PLUS SUBSCRIPTION
For a limited time, we're offering 20% off a New Atlas Plus subscription.
Just use the promo code APRIL at checkout.BUY NOW
However, Mehrabian himself has pointed out that his experiments dealt with communications of feelings and attitudes, both things that robots don't have – yet. Nevertheless, a lot of communication is non-verbal and the researchers at Georgia Tech say that developing robots that move in a more human-like fashion will allow them to better communicate with people and allow people to better understand how to approach them and how to interact with them.
To prove their theory the researchers programmed a robot called Simon to perform a series of human movements taken in a motion-capture lab. Instead of the usual jerky motion, they optimized the movements to allow more of Simon's joints to move at the same time and for the movements to flow into each other instead of being carried out in a start/stop sequence. The researchers then asked some human subjects to identify the movements Simon made.
"When the motion was more human-like, human beings were able to watch the motion and perceive what the robot was doing more easily," said Ph.D. student Michael Gielniak, who carried out the study with Andrea Thomaz, assistant professor in the School of Interactive Computing at Georgia Tech's College of Computing.
The researchers also tested the algorithm they used to create the optimized motion by asking the human subjects to copy the movements they saw Simon making. They found that the subjects had a much easier time mimicking the human-like movements.
"We found that this optimization we do to create more life-like motion allows people to identify the motion more easily and mimic it more exactly," said Thomaz.
Thomas and Gielniak presented their findings at the Human-Robot Interaction conference in Lausanne, Switzerland this week and now plan on looking at how to get Simon to perform the same movements in various ways.
"So, instead of having the robot move the exact same way every single time you want the robot to perform a similar action like waving, you always want to see a different wave so that people forget that this is a robot they're interacting with," said Gielniak.