Prof. Dr. Rober Riener
Full Professor at the Department of Health Sciences and Technology, ETH Zurich

Key takeaways

  • AI and machine-learning technologies are used to create devices and innovations that can help individuals overcome physical injuries and disabilities.
  • The development of wearable systems, attempts to make assistive devices intelligent enough to know how humans move, and employing real-time intelligence to predict the patients’ next moves are examples of the most important trends in rehabilitation technology today.
  • ArmeoPower is the first exoskeleton device to support the intensive training and rehabilitation of a paralysed arm.
  • By applying evidence-based methods rather than experience-based methods, therapists can ensure better outcomes for patients.

The potential of artificial intelligence (AI) to deliver groundbreaking benefits is not limited to the digital realm. AI and machine-learning technologies are increasingly being used by researchers to create devices and innovations that can help individuals overcome physical injuries and disabilities. Robert Riener is a professor of sensory-motor systems at the Department of Health Sciences and Technology at Swiss public university ETH Zurich. Professor Riener conducts research into neuro-prosthetics and has developed robots and interaction methods for motor learning in rehabilitation and sports. More than a decade ago, he developed the arm exoskeleton ArmeoPower, which speeds up the post-stroke recovery process. Currently he has published more than 500 peer-reviewed journal and conference articles as well as 26 patents to his name. We spoke to Professor Riener to discuss how technology is helping people who suffer from neuromuscular injuries and physical disabilities.

World Health Organisation1 figures suggest 16% of the global population experience some form of significant disability. In which areas do you think developments in bio-mechatronics and informatics/AI are likely to have the biggest impact?

We are currently seeing great improvements in the development of single components in rehabilitation technology and robotics. This includes better batteries and better motors capable of producing more force with less energy. At the same time, we also have better computational power – new algorithms as well as deep-learning and other machine-learning technologies. The challenge now is to take these great components and combine them with complex technical systems that are capable of producing better technical assistance – for example, supporting people with disabilities and helping with the rehabilitation process in clinical environments.

One of the most important trends in the field of rehabilitation technology is the development of systems that are wearable, so people can carry around the technology they require for support. But this requires mobile components that are lightweight and not bulky, and which do not need to be plugged into a power socket with a cable. We need batteries that have a sufficient level of capacity.

What role does the neuromusculoskeletal system play in how humans move and interact with the world – and what happens when elements of the system fail?

It is a complex system that allows our bodies to perform movements or to communicate – to speak and listen. We need our eyes and the receptors in our muscles, tendons and joints. We also need the vestibular organ in our ears to keep our balance, and we need our actuator system – the muscles themselves. Then, to combine the sensory information from our receptors with the muscles, we need some kind of motion intelligence: our brain and our experience or memory. However, if any of these components is lacking or malfunctioning then it will not be possible to carry out successful, functional, and meaningful movements.

How does this understanding of how people perform and control movements help prevent injury or restore lost motor function?

We need to understand how a healthy person moves in order to design an apparatus that can compensate for lost movement functions. This is where assistive devices come in. Not only do we need to make the devices intelligent enough to know how humans move, but we also need real-time intelligence. What I mean by that is that the devices can comprehend what the person is trying to do – whether it is to stand up, walk, or speak – and provide the appropriate support during that moment. Therefore, we need effective sensors and a high level of machine intelligence.

Not only do we need to make the devices intelligent enough to know how humans move, but we also need real-time intelligence.

We have other projects in our lab where we monitor the behaviour and physiological functions of patients all day long over several days. We monitor, for example, how long they sit in a wheelchair, how long they sleep, and when they eat. We measure how long they watch television or take part in some kind of para-sports while also recording heart and respiration functions. The aim of recording these activities is to better understand the risk these people might have of developing certain symptoms.

For example, if people are spending too long sitting or not doing enough exercise, they might develop decubitus ulcers – pressure sores on the buttocks – which are very difficult to treat. Alternatively, they may develop an imbalance in the cardiovascular system that can result in very high and life-threatening blood pressure spikes. By monitoring a patient's behaviour and physiology, we are sometimes able to detect these problems before the patients themselves do, or their doctors – and by predicting a problem at an earlier stage, we have a better chance of preventing it.

You helped introduce ArmeoPower, the world’s first commercially available robotic arm exoskeleton for neurorehabilitation. How did it revolutionise arm and hand therapy, and what were the key challenges you faced during the development process?

ArmeoPower was the first exoskeleton device to support the intensive training and rehabilitation of a paralysed arm – it is rather common in stroke patients. It allows for both a greater number and a faster rate of repetitions than a human therapist can manage. Rather than replacing the therapist, ArmeoPower is a tool that helps them carry out the training faster and more intensively, which leads to better outcomes for patients.

Rather than replacing the therapist, ArmeoPower is a tool that helps them carry out the training faster and more intensively.

Another important aspect of the design is that we made the robot responsive: it can understand how much muscle force the patient is able to contribute to a certain movement. The robot then provides only the minimum level of support required. The thinking here is that if the robot did everything, the patients would get lazy, and the learning progression would be small. We also developed gamification strategies, utilizing games and virtual reality to increase the motivation to complete the training – which can be very exhausting and boring.

Neurorehabilitation has a vital role to play in patient care. But how can neurorehabilitation specialists ensure optimal patient outcomes?

It is important that therapists apply evidence-based rather than experience-based methods. Experience-based means the knowledge the therapist has is isolated and subjective. With evidence-based methods, we have a level of confidence that the devices, or tools being used in a particular therapy, are beneficial and functional not just for individual patients but for the majority of patients. Evidence-based methods are already common in pharmaceutical industries where medications are tested with thousands of patients.

It is also important that scientists speak to therapists to inform them about innovations and potential therapeutic improvements. But, at the same time, we should listen to therapists to understand the problems they are facing with the patients and their own, often exhausting, work. We need a dialogue between the needs of the patients, as expressed by the therapists, and the technical methods. Only in this way can we find the most appropriate and effective solutions. Often, engineers think they know the solutions while therapists think they can solve problems using the methods they have always used. However, there needs to be an openness to new ideas, and this requires both sides to talk before embarking on any new project or innovation.

Cybathlon, ETH Zurich’s unique non-profit project which challenges teams to develop assistive technologies, will take place in Kloten, Switzerland in October (as well as in hubs all around the world). What led you to set up Cybathlon?

My initial inspiration for Cybathlon was that today’s assistive technologies are not functional enough, and therefore also not well accepted. For example, wheelchairs are usually unable to climb up stairs, over curbs, or even go over uneven terrain in the forest. It is still difficult to connect a prosthesis to a leg or an arm, while powered prostheses are complicated: their batteries do not have enough capacity and need to be recharged after perhaps four hours of use. Prostheses for legs are usually not powered, but if you want to go upstairs or up a slope, you do need some power to provide the required energy while also ensuring an efficient and symmetrical gait.

So, there are a lot of improvements needed in this area, and that is what gave me the idea for Cybathlon – to animate people and inspire teams to develop technology that functions better. The event is not like the Paralympics where the athletes have to run a certain distance in the shortest time or lift the heaviest weights. We have real-life challenges, for example, to use an arm prosthesis to open a jar of marmalade, then cut a loaf of bread and butter it.

What was your original inspiration for working in robotics?

I often get these questions from journalists, asking if there had been a tragedy in my family – but fortunately, that was not the case. I was probably triggered by a fascination for technology. My father was a car mechanic in Munich and, at the same time, I always loved medicine and the human body. So I merged the two topics. Firstly, studying mechanical engineering and then stepping into the field of medicine and rehabilitation. My aim was to make others happy – not just by helping patients to be more mobile, but also by setting up events or developing aspects of our research that could entertain people.

About the author

Robert Riener is Full Professor for Sensory-Motor Systems at the Department of Health Sciences and Technology (D-HEST), ETH Zurich. He has been Assistant Professor for Rehabilitation Engineering at ETH Zurich since May 2003. In June 2006 he was promoted to the rank of an Associate Professor and in June 2010 to the rank of a Full Professor. He is also active as full professor of medicine in the Spinal Cord Injury Center of the Balgrist University Hospital (University of Zurich). Robert Riener studied mechanical engineering at TU München and University of Maryland, USA. He received the Dipl.-Ing. degree and the Dr. degree from the TU München in 1993 and 1997, respectively. Riener develops robots and interaction methods for motor learning in rehabilitation and sports. He is the initiator of the Cybathlon, which was honored with several international awards. In 2018, Riener obtained the honorary doctoral degree from the University of Basel.

Related insights

Contact us

Make an inquiry

Fill in an inquiry form and leave your details – we’ll be back in touch.

Whether you have a question or a request, we will be happy to get in touch with you.

Introducing our leadership team

Meet the members of the team responsible for UBS Asset Management’s strategic direction.

Find our offices

We’re closer than you think, find out here.