Future co-working: a space where humans and robots will learn from each other? | Science| In-depth reporting on science and technology | DW | 26.04.2016
  1. Inhalt
  2. Navigation
  3. Weitere Inhalte
  4. Metanavigation
  5. Suche
  6. Choose from 30 Languages


Future co-working: a space where humans and robots will learn from each other?

The fear is that robots will replace humans in the workplace. But engineers at Technical University Dresden think robots and humans can collaborate - via wearable technology for humans to train robots.

DW: How does your technology work?

Jan Falkenberg: We have intelligent clothes and with them we can control LBR ("Leichtbauroboter" - robots used in lightweight design and construction) made by Kuka, and record the motions. LBRs are not as expensive as the big robots, so SMEs can afford them. But the problem is that they don't know how to program them. So we're developing intuitive teaching methods and simple workflows. You can teach the robots to do complex workflows and also repeat them.

So it's not that humans control the robots. You want to teach the robots by showing them the movements and processes first...

Exactly. We teach the robot through intuitive learning, using intelligent clothes, and then we automate the process.

And you've got electronics in your jacket and you're wearing a glove with sensors. Tell us about that.

The glove is connected via Bluetooth to the intelligent jacket and the jacket communicates via WiFi with the robot, and we also have haptic feedback in the glove - when I grip, I feel the grip - so that we have a flow of intuitive learning to the robot. There's a little microprocessor in the glove, and that's how we collect data and calculate it at our base station.

There's been a lot of talk in recent years about robots replacing humans in factories and in other workplaces. Is this one area where humans and robots can work together?

Deutschland Jan Falkenberg auf der Messe Hannover

The intelligent jacket is fitted with sensors in one sleeve and a pocket full of electronics.

Yes, this is a main focus of ours. Once we've taught the robot, we can enable very simple co-working, because the robot is aware of the human nearby. [eds. note: The robot senses the human via sensors in the jacket.] And because of this the robot can adjust its movements, because it knows, "Ah, there's a tall person, wearing a smart jacket." And the human can interact with the robot by making simple gestures and the gestures are recognized, and the robot can continue his work because it recognizes the gestures. This is step two after teaching and refining the processes, giving us a simple connection point for interaction between the human and the robot. It enables robot co-working.

You're pretty optimistic about the future with robots. Aren't you worried about robots taking over your job? Robots could even replace programmers and engineers.

Yes, maybe. But we're software engineers so we're happy about that, because we can build complex software systems to enable these features. This robot has a new API [application programming interface] from Kuka, which enables us, as software engineers and not robotics guys, to do new, crazy stuff with the robots.

Deutschland Jan Falkenberg auf der Messe Hannover

Jan Falkenberg and his team are among others developing industrial systems for the 2016 Kuka Innovation Award

So let me get this straight. You're really not worried that robotics, and things like artificial intelligence and automation will make you redundant one day?

Perhaps… But there are still so many difficult processes we can teach robots. So I think we've got another 10 or 20 years to go before we become irrelevant!

Jan Falkenberg is a research assistant at Technical University Dresden. He and his team are among other finalists at the 2016 Hanover Fair, presenting their work for the 2016 Kuka Innovation Award. This year's task is titled the "Flexible Manufacturing Challenge." The prize is awarded at the industrial technology fair.

DW recommends