Robots learn human movements

Though many tasks are being taken over by robots, this is also creating numerous new jobs and business ideas. In Dresden, for example, six young scientists from the start-up “Wandelbots” are working on a solution which could soon revolutionise companies’ production processes well beyond Saxony’s borders. The idea is to use smart clothing to teach robots sequences of human movement, so that they can then carry out tasks on their own. When a worker lifts his arm and grasps an object, the robot can immediately trace his movement: the data is transmitted to a software program via the person’s jacket and gloves. Within just a few minutes, the machine thus knows exactly what to do, and can also adjust rapidly to new requirements. Thanks to the Wandelbots, tasks which specialists and programmers until recently needed several days to achieve require very little time or expert knowledge. Science fiction? Not at all! Various companies from all over Germany are helping the Wandelbots start-up get their solution on the market as soon as possible. One Berlin investor is even investing a six-figure sum in the project. TU Dresden and the state of Saxony are also lending a hand to ensure that Saxony’s companies can soon benefit from the new technology. We visited the team and learned what was behind the project from one of its founders, Christian Piechnick.

"We received enquiries from an incredible number of different sectors. So we said, okay, let’s start a company."

Christian, your start-up helps companies program robots and teach them new work steps in no time at all. Why is that important for many companies?

Large companies, especially, are noticing that more and more start-ups are popping up which are able to offer individual products. Collectively, that is taking quite a chunk out of their clientèle. As a result, they have to develop processes to make their huge production plants work more flexibly. In addition, there are a lot of dangerous, unpleasant, monotonous and uncreative jobs that people cannot or do not want to take on – for example in chemical production processes, when things that have come into contact with chemicals have to be cleaned in acid. In such cases, the robots we program help lessen the workload and increase workers’ safety.

What was the moment when you realised that your project for programming robots using smart clothing could work as a business idea?

Two and a half years ago, when we were young scientists, we presented our solution at the Hannover trade fair, and came home with a huge pile of business cards. Everyone asked when they would be able to buy it. There was a confectionery manufacturer that needed robots to sort chocolates into chocolate boxes, or a film team from the USA that wanted to use the jacket to remotely control robots with cameras on them during car chases. We got enquiries from an incredible number of different sectors. So we said, okay, let’s start a company.

What does the smart clothing look like that the people wear to instruct the robots?

We use standard items of clothing – jackets, gloves, sweaters – and integrate sensors and actuators into them. The sensors record data on people’s body movements; the actuators give the wearer feedback, for instance by vibrating. When you teach the robot to grasp something, it’s important that the clothing, such as the glove, gives you the feeling that you are grasping something. Most of our start-up's value, about 98 per cent, comes from the software we've developed. While something is being demonstrated for the robot, our software records all the data it can collect. This data is analysed, and an automation process is created in the form of a script. The more frequently I teach the robot something, the bigger the dataset and the more accurate the robot's movements.

What sets you apart from other companies offering similar solutions? Is there a lot of competition in your industry?

No, there really isn’t! There are a handful of start-ups around the world that are moving in a similar direction. We have very good connections to some of them. There are also research groups testing out a similar approach. But as far as I know, we are the only actual company taking this particular route of programming robots via smart clothing. The topic of software for robots has been totally neglected for decades. Of course, there are thousands of companies programming robots. Their business model is to help companies with automation. You want a step in your workflow automated? Then these providers buy robots and build a manufacturing cell around them, program everything, sort out the safety technology and run it all. If you want to change something, it’s complex and incredibly expensive.

You are all from TU Dresden, but it was by no means the only possible course for you to found your company here. Why did you choose to stay in Saxony?

We have a Berlin investor who wanted us to come to Berlin. We are supported by business angels in southern Germany who want us to come to Munich or Stuttgart. But the advantage Dresden has to offer is that this is the biggest microelectronics location in Europe. That means there are a lot of firms here that revolve around automation. And those are precisely the kind of companies we want to work with. We also receive support from the business hub, the state chancellery and the city. Moreover, it is easier to gain access to people at higher management levels in companies than in many big cities. Then there’s the fact that we all have roots, and plenty of support, here. We want to stay here, and believe that a huge number of Saxon companies can really make good use of our solution.

Finally, a personal question: should we be worried that we’ll be replaced by robots one day?

There’s a long way to go before robots can proactively and creatively solve problems. Robots won’t be able to do the things that set us humans apart even in 10 or 15 years. Robots are pretty good at learning repetitive things that you tell them to learn, but as soon as something is different, or something unexpected happens, a human has to deal with it. Whenever robots are used, there are also people who have to monitor them.

When the steam engine was invented, lots of people thought that it would put people out of a job. But it didn’t, and it’ll be the same with automation.

Diese Erfolgsgeschichte finden Sie auch auf so-geht-sä
Bildnachweis: Anne Schwerin