Rehabilitation patients require a reliable relationship with their physical therapists. But what if that therapist is a robot?
Although not a mainstream technology in clinical settings (…yet), socially assistive robots (SARs) are already being used as effective rehab strategies for patients who are recovering from strokes or other diseases causing severe functional deficits.
For the SAR technology to become commonplace, however, robotics engineers will need to build both technology improvements, as well something slightly more abstract and complicated: Trust.
In the Science Robotics journal this month, researchers concluded that the development of “trustworthy" socially assistive robots will require a number of advancements in predicting the machines’ behavior.
The team, which included neuroscientist Dr. Philipp Kellmeyer of the Freiburg University Medical Center, asserted that aspects like philosophical and developmental psychology must also be considered.
Kellmeyer spoke with Tech Briefs about the improvements necessary to make socially assistive robots valuable and dependable.
Tech Briefs: What kinds of improvements are needed to make SARs more trustworthy to a patient?
Dr. Philipp Kellmeyer: In the paper, we argue that important prerequisites for a trustworthy interaction between humans and SARs are:
- Safety of the robot's behavior
- Shared intentionality/predictability of behavior
- Mutual attunement, i.e., sensitivity of the robot toward fluctuations in the human counterpart's abilities (due to fatigue or lack of motivation, for example). The patient, however, also needs the capacity to attune to the robot's behavior.
Tech Briefs: What is being done to make the social robots safe?
Dr. Kellmeyer: Regarding safety, I think the regulators (specifically in the U.S.) and the industry/academia are quite aware of and sensitive toward the need for safety, so I think that their existing guidelines can go a long way in ensuring a safe interaction.
What is less clear, however, is whether these same regulatory frameworks and industry standards will be sufficient for ensuring safety in robots that are operated with advanced machine learning/AI systems. On the one hand, the increase in system "intelligence" would be an important condition for the robot to be able to achieve something like highly adaptive and socially appropriate behavior. On the other hand, this same advanced "intelligence" may open accountability gaps in cases of grave system errors with subsequent damage to humans.
Tech Briefs: How can your second point, the predictability of behavior, be improved?
Dr. Kellmeyer: For an SAR to be able to interpret and predict the behavior of humans, it would need to be powered by advanced methods of machine learning, or deep learning, to be able to adapt to an individual user based on learning the user's behavioral data — and any other helpful kind of data in machine-interpretable form.
So, in short: smarter SARs will likely be more trustworthy for humans to interact with — perhaps much as in the case of human-human interaction.
Tech Briefs: What improvements can be made regarding “mutual attunement?”
Dr. Kellmeyer: Perhaps giving the SAR more sensing capabilities — audio for inferring emotional states from the prosody of speech patterns, for example — will provide the SAR with valuable data for attuning to humans.
Tech Briefs: How possible are these improvements?
Dr. Kellmeyer: Given the recent advances in machine learning and big data driven analytics, I would submit that there is a lot of room for improvement in all those domains.
Tech Briefs: Will there always be trust issues between patient and robot?
Dr. Kellmeyer: No. It will depend on the mutual "vibe" between the patient and the robot — again, much like in human-human interactions, which also have the unfortunate tendency to go awry or fall apart because of misunderstandings or for petty reasons.
Tech Briefs: In what medical scenarios, do you envision SARs being ideal?
Dr. Kellmeyer: As a clinician, I am certainly inclined — some would say biased — to take a patient-centered view, to make the ideal scenario one in which the benefit for the patient is maximized. As these SARs have so many potential applications — rehabilitation, care, treatment of psychiatric disorders — this would, of course, need to be discussed in terms of the specific context.
Examples that come to mind would be: A patient with a severe motor deficit needs high-intensity motor rehabilitation based on highly strenuous and repetitive (and thus demotivating) movement exercises. This patient lives in a rural area where not enough highly skilled physical therapists are available to administer high-intensity physical therapy. Initially, the qualified physical therapist works a few sessions alongside the SAR with the patient, so that the patient builds a trusting relationship. Later, the SAR can take over the repetitive and strenuous part of the actual exercises and can either give motivational input, or the PT could join 1 or 2 sessions per week and provide the patient with motivation and encouragement.
Tech Briefs: What kind of future do you envision with social robots? Will human-robot interaction be commonplace, do you think?
Dr. Kellmeyer: This depends on the time scale we are talking about. For the next decade, my guess is that SARs will be mainly used in specialized medical facilities (stroke rehabilitation centers, for example). But if the technical progress remains steady, they might also enter the consumer market and be made available to more people.
The collaborative research team included neuroscientist Dr. Philipp Kellmeyer of the Freiburg University Medical Center; Prof. Dr. Oliver Müller from the Department of Philosophy of the University of Freiburg; Ronit Feingold-Polak; and Prof. Dr. Shelly Levy-Tzedek from the Recanati School for Community Health Professions at the Ben-Gurion University of the Negev, Beer-Sheva, in Israel.
What do you think? What kind of future do you envision with social robots? Share your comments below.