Q & A with Robotics Designer Kathleen O’Donnell

TEDMED: You mentioned in your Talk that as you became more involved in medical robotics, you realized there are many non-traditional approaches to robotics. The exosuit, a soft wearable robot being a great example of a non-traditional robot. Are there any other innovative designs in the field of medical robotics that stand out to you, or that you have worked on recently?

Kathleen O’Donnell: There are tons of interesting robotics approaches out there! I recently read an article in Nature about a system that consists of individual robotic elements that come together and behave cooperatively to achieve locomotion and other complex tasks, similar to the way that the cells of living organisms work together to achieve complex functional behaviors. This example really helps to highlight the massive diversity in robotic approaches today.

TM: New technologies often require a lot of time and money to create. Are these things the biggest barriers to creating new innovations in medical robotics?

KO: One reason that it takes so much time and money to create new innovations in medical robotics is that designs need to be iteratively tested in representative use scenarios to properly develop and validate the designs. Then you still have to begin your summative clinical and engineering testing to ensure that everything is performing safely, effectively, and in compliance with relevant regulatory agencies (such as the FDA). One thing that has really helped to accelerate the development of the exosuit is that from the very early stages, we always were able to involve stroke patients and physical therapists in the device testing, through an IRB-approved protocol. This “early and often” approach to testing with actual users of the device (both the therapists and the patients) helped to ensure that each iteration of development was helping to move us closer to our end goals and allowed us to course-correct before we got too far off track.

TM: The exosuit was adapted to address mobility issues stemming from neurological disease. Do you think soft wearable robots like the exosuit will be used in a more widespread way in the future?

KO: Absolutely! The Exosuit for Stroke Rehabilitation that I discussed during my Talk recently achieved several major milestones, including completion of a clinical trial and achievement of FDA clearance and CE marking, meaning that the exosuit is now commercially available for clinics in the US and Europe to purchase for use in their stroke rehabilitation programs, making this the first (of many) widespread clinical applications for soft exosuits. Furthermore, the technology which comprises the core functionality of the soft exosuit is essentially a platform technology that can be adapted to a wide variety of applications. By leveraging the knowledge gained from developing the exosuit for stroke rehabilitation, we can more quickly develop systems to support additional joints, such as the hip or the knee, as well as additional patient populations, such as MS, Parkinson Disease, or TBI, for example. It’s really exciting to see how the first exosuits have lead to such a robust pipeline of innovation.

Photo Credit: Wyss Institute

TM: In your Talk, you placed a great emphasis on the fact that the focus is always the people the technology is helping, do you think your experience as a patient plays a part in this mindset?

KO: I think my experience as a patient has certainly helped me to empathize with the patients we work with, and to understand why walking ability is such a powerful component of patients’ quality of life. However, even without this experience, I think it would be impossible to work as closely as we do with patients and therapists and not develop a deep sense of empathy and understanding for the challenges they encounter on a daily basis. The teams I have worked on have always placed an emphasis on going the extra mile to “get out of the lab” and better understand the people who are using these robots and understand what they are trying to achieve, and it is this mindset which continues to be instrumental to informing the design of exosuits throughout their evolution.

Massive Science on Kathleen O’Donnell

Massive Science is a digital science media publication that brings together scientists and the science-curious public. The team at Massive joined us at TEDMED 2018 and covered talks by various speakers including Kathleen O’Donnell. Check out their coverage of O’Donnell’s 2018 TEDMED Talk below.


Photo Courtesy of Massive Science

There’s been a bevy of heavy metal, superpower-imbuing robotic suits in pop culture — think Halo, Avatar, or Iron Man. In fact, these fictional portrayals were what inspired researchers at Harvard’s Wyss Institute in the Biodesign Lab to develop a new exosuit.

Initially, the goal of the exosuit project was to develop military applications. (Not surprising, considering the project was funded primarily through DARPA.) The researchers combined traditional robotics with flexible fabrics and lightweight parts, resulting in a soft, wearable design.

The scientists realized this technology could also be a medical tool. Kathleen O’Donnell, a staff industrial designer at Harvard’s Wyss Institute, met with clinicians and quickly honed in on stroke patients, who often suffer from weakness and loss of control in one side of their body. O’Donnell’s team envisioned designing a suit that could be attached around the waist and calf, to help stroke patients balance their strides, reducing the effort it takes to walk. Volunteers were soon recruited as study participants, and a team of roboticists, industrial designers, control engineers, and physical therapists began designing, testing, and iterating the suit.

The team quickly faced several major challenges. “We have algorithms that measure the way you walk and try to predict when are you taking a step so that we can time the assistance,” explains O’Donnell. This kind of responsive assistance was easy to control in soldiers, since they tend to walk with symmetric, regularly-paced strides. But stroke patients tend to walk with different compensations and irregularities.

“Her foot looked so much more confident, so much more stable. She was able to stand up straighter.”

“Everybody walks a little bit differently after their stroke. They have different compensations they may use. One person might hike their hip up as they’re walking. One person may swing their leg around as they’re walking,” says O’Donnell. “We had to understand how to ignore [the compensations] to some extent, but still get the information that we needed about their gait to time the assistance with their particular gait pattern.” This personalized capability required the team to build adaptable algorithms that adjust the suit’s required assistance with every step. The resulting exosuit never imposes how to walk — it just helps the patient walk naturally.

Another major difference between soldiers and stroke patients is body type. While it’s easier to design for the typically fit physiques of soldiers, stroke patients’ physiques vary widely. Since the suits need to attach closely to a patient’s body, individual body types can significantly change the design of the suit. O’Donnell explains, “From an apparel design side, understanding both the range and mechanisms we were using to attach [the exosuit] as securely as possible to the patients became more challenging.”

With a diverse group of patients, the team built a toolbox of strategies to individually fit an exosuit to every user. During one testing and recording with a patient, O’Donnell describes the patient’s transformation as dramatic. “Her foot looked so much more confident, so much more stable. She was able to stand up straighter.” While she acknowledges that fitting the suit required time, even without any optimization, the change in patients was frequently instantaneous.

From the beginning, O’Donnell and her team focused on patient volunteers who had experienced strokes and could immediately benefit from the exosuit. “It has been such an amazing process to work with all these volunteers from the community,” says O’Donnell. “Our first volunteer is still one of the volunteers who comes in, five years later.” Licensing the technology from the Wyss Institute, O’Donnell guided the transition of the exosuit and began to manage clinical trials in the hopes of making the suit available to the millions of stroke patients in the United States today.

“We are starting in stroke, but we could potentially see suits for MS or suits for Parkinson’s.”

So far, the exosuit has been tested on more than 40 patients. Of course, there will be potential challenges in scaling the technology. “We have made as much of an effort as possible to get as diverse a range of patients as we can. That includes body sizes and types, walking speeds, [and the] types of assistive devices they use,” O’Donnell says.

Giving freedom back to stroke patients is just the beginning. O’Donnell says the exosuit could help many other kinds of patients too. Other injuries or disorders are also on their minds. “We are starting in stroke, but we could potentially see suits for MS [multiple sclerosis] or suits for Parkinson’s.” With the ability to quickly alter and control the assistance, the exosuits could help people undergoing physical therapy by providing assistance when needed and taking it away to help rebuild strength. On the other side of the spectrum, the exosuit could be used at home to provide general, consistent assistance. Luckily, being made out of fabric helps reduce the overall cost of the exosuit. The possibilities for exosuits in medicine will be exciting to watch.


About the author: Joshua Peters is a PhD student in Biological Engineering at MIT. Around two billion people in the world are infected with a microscopic bug called Mycobacterium Tuberculosis. Despite this, only a fraction develop tuberculosis. And a fraction of those infected – almost 5,000 a day – die. He puts on Stranger Things-esque protection equipment and probes these bacteria to ask, what allows them bacteria to win this tug-of-war? To understand this variation, he looks at how both human and bacteria cells change on a genetic level in response to each other, as a member of the Blainey Lab, located in the Broad Institute, and Bryson Lab, located in the Ragon Institute and MIT.