Please give a warm welcome to our new perceptual robotic friends

June 22, 2009

by Mark Ollig

In the future, robots will not only carry out the assigned tasks we humans expect from them, they will come with “built-in” intuitive abilities to anticipate our intentions and make suggestions to us.

This capability to anticipate our actions could make human-robot relations seem more normal. Heck, it is hard enough for me to get along with humans who seem to enjoy pointing out all my infallibilities, how will I explain myself to a walking, talking, and thinking robot?

I would not worry about these futuristic robots pointing out our mistakes or telling us how to correctly change a tire – at least not just yet.

From the information I have been reading, it seems some research groups want to construct robots which will act more like “companions” than “workers.”

But in order to be a companion, they will need some ability to “learn” to adapt to our human interactions in a natural and “non-robotic” mannerism (does this make sense? I thought it did when I wrote it).

Do not be alarmed just yet; consider that today’s most intelligent robots still can only perform what we intellectually-superior-brained humans program them to do.

But in order to play a more pro-active part in our daily lives, we humans will need to figure out how to create robots with some complex decision-making and logical task accomplishing abilities.

You know what that means, right?

The new robots of the future will need to be asking us questions in order for them to learn so they are able to decide a course of action to best complete a task or anticipate what is expected of them to do.

Are scientists actually working on this? Why, of course they are.

Let’s take a look at the effort being done in this very field by the folks at the European Union-funded “Joint-Action Science and Technology” or “JAST” project.

JAST has what it calls a “multidisciplinary team,” which is discovering ways a robot can anticipate the actions and intent of its human partner as human and robot go about working “collaboratively” on solving problems or completing a job.

“You cannot make human-robot interaction more natural unless you understand what ‘natural’ actually means,” says JAST.

JAST confirms the fact that hardly any examinations have analyzed what the underlying creative processes are which people – when working together – use in order to accomplish a shared objective.

One of the projects JAST conducted was studies of human-to-human relationships and general observations which they think could be incorporated into the advancement of robotic behavior to make them react in a more natural way.

So, not only does JAST want to create a competent robot to assist humans in completing tasks, they want to build one with a set of human brain-like neurons to “mimic” human thought when observing an activity.

JAST said they discovered during shared work assignments, people study their partners as the observer brain copies the observed action and tries to make sense of it. JAST went on to say “the human brain mirrors what the other person is doing either for motor-simulation purposes or to select the most adequate complementary action.”

The JAST robotics folks have put together a method which integrates this ability for observation and emulation.

“In our experiments the robot is not observing to learn a task,” explains Wolfram Erlhagen from the University of Minho and one of the project consortium’s research partners. “The JAST robots already know the task, but they observe behavior, map it against the task, and quickly learn to anticipate [partner actions] or spot errors when the partner does not follow the correct or expected procedure.”

“Our tests were to see whether the human and robot could coordinate their work,” Erlhagen went on to say, “Would the robot know what to do next without being told?”

How is the JAST system different to other experimental robotic programs?

“Our robot has a neural architecture that mimics the resonance processing that our human studies showed take place during joint actions,” says Erlhagen. “The link between the human psychology, experimentation, and the robotics is very close. Joint action has not been addressed by other robotics projects, which may have developed ways to predict motor movements, but not decisions or intentions. JAST deals with prediction at a much higher cognitive level.” Erlhagen concluded.

Gosh, if this doesn’t remind me of an original Star Trek episode.

Only, if I remember it right, it didn’t work out too well for the robots because they studied the human characteristics of one Harcourt Fenton Mudd.

JAST is located at http://www.euprojects-jast.net.

This week’s “Web Site of The Week” features the European ICT (Information and Communication Technologies) web site.