David Appink wants to make robots that make people happy

Most of the robots I’ve built are quickly gathering dust. I usually don’t even mind. But I think now: there is a huge problem in the labor market, people are escaping from many essential occupations, and work really needs to become more attractive. Bots can help with that.” But then people have to use them.

David Abbek would do things very differently.

Appenk, 44, has been a professor of human-robot interaction at TU Delft since 2018. He got his spurs in control of smart vehicles. According to him, the future of driving is not the fully self-driving car, but the people and cars who share the responsibility for driving, such as a rider and his horse. He built a research group on haptic feedback, human-robot tactile interaction, but now he will lead a consortium to develop workplace robotics. This categorically includes not only engineers, but also psychologists and ethnographers. He believes technicians need to understand the workplace much better, not just swing robots over a fence.

Bots don’t think much about how they actually use them

Many come into the workplace. Technical, with productivity and all kinds of measurable things, social, because work is also a place for social contact, for self-development and learning skills. When the Corona crisis happened, it suddenly became clear what professions really matter. Many of these jobs are bodily, and in many ways undervalued. The employees are leaving. “People come to him with a question, Can’t we solve this labor shortage with robots?

“David 10 years ago he would say, ‘We’ve already done that. I’ve often worked in multidisciplinary teams on robot support. For example, we made a very nice lifting robot for health care staff. Then we talked to health care workers. Too big and heavy and they didn’t use it. I thought I did my best. If you want a lifting robot, this is what I came up with.”

The position of many robotics scientists is Apink’s experiment. “We don’t think much about how to actually use it.”

man and machine

This has not been a problem for a long time, robots mainly operate in closed environments, at a distance from people. Now robots appear in the human world. So you really have to understand the impact of technology. But I never learned that.”

However, this approach was not always self-evident, Apink realized as he prepared for his inaugural lecture as a professor. “I was then reading about cybernetics, a term coined after World War II by Norbert Wiener, one of the founders of the field of control engineering. The word cyborg is derived from it, it is about the relationship between man and machine. It turns out that Wiener was working with Gregory Bateson, an anthropologist The wonderful. In cybernetics, the scientific disciplines merged, but that was completely lost.”

“Then I said, ‘My goal as a professor would not be to specialize more in human-robot tactile interaction, but to connect people and robots. I took it very seriously – I even married an anthropologist – and read and learned a lot about how psychologists and anthropologists do science.”

What began as a fun intellectual challenge – a “scientific feast” – to unite disciplines, has turned into an enthusiasm for solving problems in the labor market.

Robots are not used for people in the workplace, they do not make work more attractive

Apink stresses that healthcare is a clear example of the incompatibility between robots and humans, but it applies to many sectors. „Take the pick of orders, Amazon-like companies. Robots do a lot in the process, so some consider it a true ideal sector. But robots can’t do everything and what’s left is a corrupt business.”

The result: people run away. The solution: more bots. In the end with dire consequences. “The workers there are wearing diapers! Everything is optimized for robots and they will never have to go to the toilet. The fact that people are falling is not an issue by those companies. The profit margins are excellent. They pull someone new off the street because you don’t have to be able to do much. In any case “.

“The idea often is that robots should help people with tasks that are boring, sloppy or dangerous to be. Who wouldn’t want that? But in the meantime, robots aren’t used for people on the work floor, and they don’t make work more attractive.”

And Apink thinks this is possible. By knowing what makes people happy and where they fail. Then he invented robots that support, without destroying other things.

More complicated than it seems

This task is more complicated than it seems at first glance. “A while ago I was in a company making gears. There they wanted to make it easier to work with robots. Previously concerned researchers asked precisely what people did not like about their work. Sometimes it was annoying to have to pick up tools, the answer was. Then someone was spending time fun and then had to walk somewhere again. It also took a while.” The innovative solution was a robot cart carrying tools. But walking to the instrument in the meantime also caused someone to have a conversation with a colleague, seeing that stock was running out as they passed. “You guessed it, productivity went down. People found their work one-sided and unpaid.”

“Discovering such things requires more than a questionnaire, and that is the highest science. With qualitative research methods, you don’t usually have to reach out to robotics experts, until they realize what you cannot measure with locations and forces.”

From start to finish

Abbink uses the term interdisciplinary work many times. This form of collaboration goes beyond interdisciplinary work, where several disciplines are working on the same problem, but where the other – a “project leader or something” – deals with integration. He wants to collaborate from start to finish, with academics from different fields and with the workplace – that’s why Acrossdisciplinary.

“I don’t mean we listen to the pros once and then make something, but we also think and put together. Then you can see what’s going wrong in the robot cart gearbox, but it’s easy to tell. Nobody came to that, not even the people on the work floor They were very happy at first.”

Merlin Domirnick’s photo

This intense collaboration takes time. “I’m still learning a lot about how to do interdisciplinary research, and I don’t publish anything for a while. I find that exciting, because not publishing will ultimately kill an academic career.”

Fortunately, there is support from the university. Abing has set up the Fraim Research Center under the banner of TU Delft, a physical place where he can begin this way of research, and can submit a Gravitation proposal. Gravity is the largest amount of support for the NWO, with researchers given €25 million and ten years to open up a new field. Abbink writes this proposal with about 25 robotics, designers, ethicists, work and organizational psychologists and ethnologists from TU Delft, TU Eindhoven, University of Twente, Erasmus University and Radboud University.

But what does their research look like in practice? “We do basic research in robotics and quickly make prototypes, testing them on the spot with specialists. At the same time, sociologists observe and measure: How does working with a robot change, is it attractive and productive? We do video analytics, or when it comes to healthcare, For example, we’re not testing with real patients but with actors, all the things that were previously alien to robots.”

jet engine

To experience what it means, we walk into a room below. There is a robotic arm with polishing tools on hand. Below, a rotating blade from an aircraft jet engine is installed in a vise for automated processing. The robot’s arm can deliver all kinds of forces and movements, can be pushed away and can learn from corrections.

“At KLM they service jet engines. Repairs to rotary blades require a lot of experience and are hard work with risks of injury. A robot can help, but automating the process does not work: the work is very variable for that. The big question is, what should it do?” Grabbing as much as possible so that the employee only has to check is an option, but this is boring work and employees lose skills. The robot can also cooperate with the worker, holding a tool to move him very precisely to points of interest, with the worker providing the appropriate force. Or the robot can generate force, where the employee determines the situation and just has to push a little.

Apink does not yet know what is the best. “If you talk to professionals, they also think physical labor is great. At the same time, you might attract other people when the work is lighter.”

Well-meaning bots also have all kinds of unexpected effects at work, which we need to figure out ASAP. So we suppress our engineering reaction to immediately make a piece of engineering that solves an isolated problem. With this robot arm, we can easily go to the workplace and try out all kinds of ideas with professionals. This way we learn faster what makes this workplace more attractive, and what does not make this workplace more attractive. New and fascinating scientific questions and ideas are already emerging from this.”

Leave a Comment