Knut Åkesson is Professor of Automation at the Department of Electrical Engineering at Chalmers and conducts research in industrial automation. He is also the program director for the Automation and Mechatronics degree program, as well as research leader in automation activities in the Wingquist Laboratory.
We talked with him about automation and why it's important. But also, about his interest in problem solving as well as the importance of education and young people's interest in science and technology.
How did you come to be a researcher?
I grew up on a farm, which meant that I helped my parents in the agricultural work and spent a lot of time thinking about how to automate various tasks. I also had a great interest in maths and computers and liked to solve problems. I therefore studied to become an engineer in Lund, and on that path it is.
I like calculations and am driven by solving challenging problems with the help of computers, algorithms and mathematics. I like to combine computer science and mathematics linked to some kind of application, so that you do something "for real". Method and application – I thrive with one foot in each area. It is important to have a mix of both basic research and application in our research area. And here at Chalmers, we have always had a good collaboration with industry, for example through the Wingquist Laboratory and the Production Area of Advance.
What does industrial automation mean?
Within the research group, we have made the following definition: "Through the integration of perception, planning, control, actuation, and verification, we create methods and tools for efficient, flexible, and safe automation systems capable of collaborating with humans."
When it comes to perception, i.e. perceiving and interpreting information from the outside world, the latest breakthrough is cameras that collect information that is processed with machine learning. This means that the subject of automation has gone from being "behind closed gates", i.e. in an isolated automatic system where no one is allowed to enter and risk getting in the way of the robot, to automatic systems that work together with humans. Next to or even together to solve a task. What we call a collaborative robot and with which we humans can interact and collaborate. The new industrial robots differ from the old ones in that they have sensors, so that they stop if they encounter obstacles. But they are still "blind".
So, what we in research are working on now are robots that can take in what is happening in the environment. And then we need data to be able to make decisions, calculate and control the robot so that it can interact in a very natural way with humans.
The planning can then refer to which robot will take which task, when it will start, which route it will take and when it will arrive at its destination. The control can then mean that the robot must give way in the right direction if an unforeseen obstacle appears and, for example, go behind the obstacle if it is a human who is walking. It is therefore a matter of adapting to what it looks like in the moment and finding the best solution in real time. Challenges can arise here as quick decisions need to be made, while there is a lot of information to take into account. And to be able to flag deviations in time. This requires a lot of verification and testing, in both physical and virtual environments. But also, from two different aspects – testing so that the robot does as it is intended to do, and dealing with deviating situations that occur less frequently.
If we are to have a competitive industry in Sweden, we need to have a high degree of automation.
Why is industrial automation important?
This is important in several respects, for example, automation provides a very consistent quality. But perhaps most of all from a competitive point of view - if we are to have a competitive industry in Sweden, we need to have a high degree of automation.
Seen from a societal perspective, we will have a labour shortage in the long term because of an ageing population. It is therefore important to have the best and most ergonomic working environment possible. So that people can do what we do best, with tasks that are as stimulating and risk-free as possible. And that we can avoid work in hazardous environments as well as heavy, risky or monotonous tasks.
Another important and relatively new area of application is in the field of rehabilitation, where we have an ongoing collaboration with Sahlgrenska. For example, it can be about rehabilitation after a hand injury where a tendon has been cut in connection with surgery. Then it is important to challenge yourself just enough so that you do not overstrain the tendon. With the help of rehabilitation robots with sensors and actuators, it will then be possible to individualize and situationally adapt the resistance in the training.
What challenges do you see?
When it comes to industrial applications of machine learning, there is not much data that reflects unusual situations. Collecting such information is both expensive and time-consuming.
It is also always a challenge to move from academic research to industrial application. Technology development is accelerating and the gap between academia and industry is narrowing as it is often important, for competitive reasons, to industrialise new algorithms and technologies quickly.
Another challenge I see is how to achieve a transition when the doctoral student has finished his or her work, so that the result doesn’t stay at the university. Here, research institutes can play an important role as a bridge for realising research. For example, there is the Fraunhofer Foundation - Chalmers Centre for Industrial Mathematics, FCC, and the Chalmers Industrial Technology Foundation, CIT.
Well, there are many societal challenges that we face, which have a major impact on society. It is important that development takes place in interaction with society at large so that technology enriches our lives and not the other way around. Technology development and society must go hand in hand.
Finally...
... Knut emphasises the importance he attaches to education. "That's why I am the director for the Automation and Mechatronics programme. He also believes that it is important to support schools' efforts to increase interest in technology and science among young people, so that they apply for these programmes in high school. That is why he is also engaged in the First Lego League, an annual international technology competition for high school students, whose regional final takes place at Chalmers in November.
