Print 

next newsletter preview the future of robotsThe recent report titled ‘Application of novel industrial scale robot automation in elderly care’ from VTT Finland raised a lot of discussions about robotics and healthy active aging. MARIO is a Horizon 2020 funded project on ‘Managing active and healthy aging with use of caring service robots’ and the research team have a keen interest in the findings and recommendations from this report. We interviewed the author Dr Hannu Lehtinen and Dr Marketta Niemela, from the ROSE project – see what we found out here on the Future of Robotics…

Question: Your report on ‘Application of novel industrial scale robot automation in elderly care’ published in June 2017, so only some months ago was discussed more than many official documents that have been prepared or  designed by committees and were supposed or expected to influence the area – though they didn’t leave any significant mark behind. What do you think was / is the biggest virtue of your report, as you can judge from your side?

Marketta: The report gathers many ongoing developments with regard to elderly care robots and related advanced technologies, which helps any reader to get to now the state of the are of the domain. For Finland, this is especially crucial as care robotics has only recently raised major attention among researchers, decision makers and care service providers. It is also very important for our organisation VTT to show that we have significant competence in this domain.

Hannu: There seems to be hope that increasing demand for elderly care can be arranged with improving processes, facilities, task management and (nurse) assisting equipment. Many industrial methods can be applied to improve care of fragile senior citizen.

Question: You mainly focus on healthcare needs of older people – would you see some positive potential also for other markets or audiences? Like young children or people with disabilities or a general audience?

Hannu: Autonomous vehicles are clearly for us all. There have been some experiments to inhabit elderly persons with handicapped and young healthy persons in the same dormitory. Social interactions evolved and seemed to improve quality of life for them all. This may be the way to go. Potentially also with the most economical benefit.

Question: As children, did you have any special affection for robots? Did they inspire you? Or you were ‘cool’ and distantiated from all the emotional aspects of humanoid robots?

Marketta: I did not fancy robots as a child in any special way. But I recall that when I was around 10 years old, my father brought a computer from his workplace to home for us kids to play games -  it was the first computer I ever saw. I was totally fascinated about it, not just playing but understanding how it works. Much later I studied computer science and psychology and finally ended up to study robots. Now I have a daughter who is 10 years old, and she asks me to bring a robot from work to home, so she can play with it…

Hannu: The humanoid robot Little Helper of Gyro Gearloose (a Disney character) attracted me to robotics well before 1975 – the year to plan my education. Amazing on-line control was done at those days at the Helsinki University of Technology – now named Aalto University - via cassette reader interface of an ASEA robot. That got me to mainly focus my studies on robotics. Humanoid robots are typically not economically feasible, because they are competing with us – humans. We have sensing, intelligence and motion skills that are very expensive to duplicate in robots.

Question: How much can a human build a relationship with a robot, same as with other humans, without us needing to govern it? For example: if I am old and have a terminal disease or only am in a psychologically bad phase, why shouldn’t I confide the robot as I would do with a friend or my physician, sharing with the robot that e.g. I feel my life is worthless or that I dream of committing suicide, without the robot needing to report my behavior and my thoughts to another human like a professional carer or a member of the family? At least, and same as you also spot in your report, this is an issue of privacy.

Marketta: Surely a human can confide different difficult issues to a robot. I don’t see a problem here but in what follows. Is there relief and feeling of understanding after talking to a robot? Could the robot provide these to people one day?

Hannu: The Paro robot shows that some of us can make a relationship with a robot. Intelligent speakers – e.g. Amazon Alexa, that we can discuss with, will help us in practical matters like calendar and reminders, which then could help establish the relationship further.
Ethics, flawless cyber security and privacy are essential in any care related systems. Otherwise, they will be banned. I would build the system so, that if automated systems detect a new medical concern, it makes a related alarm to authorised medical personnel to check the concern using noticed “proofs” as  well as traditional methods. Otherwise notifications should be made only as planned and authorised in advance e.g. to relatives and named caretakers. I assume that more and more nations will discuss and allow euthanasia in legally controlled manner.

Robots and Artificial Intelligence (AI) will be taught to take care of persons as health care professionals would do - for example to treat depression.

Question: In human behavior we tend in many cases to break the rules that are forced due to human nature (i.e. eating ice cream while having diabetes). How would you envision robots handling such deviations? Being too strict? Being a bit lenient or even ‘covering up’ a person’s deviation by not reporting fully his/her behavior to doctors and caregivers? So, what would their mission be? (helping us never die, helping us die as late as possible or helping us to increase our happiness levels) Would this choice of robot behavior affect acceptability and take up from elderly and the rest of society as well?

Marketta: The human should decide her or his behaviour, and also how much to listen to the robot’s advice. I don’t think people would like to live with robots that try to be bossy and define how to live. But if the robot is more like an advicer, motivator, or even in the role of a mentor, the good advice would be accepted. That requires building the social relationship and bonding with the robot, maybe a specific kind of emotional trust to the robot.

Hannu: Robots should act as physicians do. Ice cream is OK if insulin and glucose levels in blood are and remain within the specified limits. We are close to automatic insulin measurements and injection. For example, sports activities in the person's calendar may also modify insulin dosing or advise to take suitable dietary snack. To improve our happiness level would be a clear task for robots. Novels and the Three Laws of Robotics by Isaac Asimov are a good introduction to robot ethics and to the complexity of the matter.

Question: Given that future generations of older people might already have long digital history recorded on different services like Facebook, Twitter, Spotify, Booking, etc. Would it be acceptable for robots to use all this data about a person to improve personalization right from the first moment or should they discover this info from scratch? What would that mean for services provided by robots?

Marketta: I think utilising the digital history of a person to help a robot to better understand her or him is a great opportunity and would increase the usability and usefulness of the robot and maybe also  the person’s emotional attachment to the robot. But that should definitely be done respecting the privacy of the person. The person could have a possibility opt out using her or his digital history to teach the robot. Maybe not many persons would use that option after all.

Hannu: It is understandable that everything will be connected in the future and our preferences are known, as – and if - we have allowed collection of hundreds of thousands of data items. Robots and all other systems – like web shops, intelligent speakers and our phones – will utilize the collected data. It does not hurt us that the data collection companies make money by selling our preferences.

Question: What do you see as main problem or obstacle in the development of a market for robots in Europe and internationally? Prices that are still high? Mentality that is pushing things to happen later? Technologies that haven’t yet matured? All of the previously mentioned? Something else? …

Marketta: There are several obstacles on the way: care robots are still quite immature with regard to the delicate and various needs of elderly care. The elderly care service providers may not know what robots are available on the market. Also prices are high. It is difficult to increase productivity in care services through robots similar to e.g. manufacturing industry, so the motivation to invest expensive robots is missing. The general conception may be that care service is something only done by humans. This prevents us to see that care tasks have many possibilities to be automated, which would then free care workers’ time to develop the social and human aspects of care, as well as create new more technology-oriented jobs in care services (e.g., monitoring and operating robots in care work).

Hannu: Producing robots in large numbers – like any product or service concept - will reduce the unit prices. However, customers do not buy products in large numbers until they get great value in return on using or applying the product or the service. It is difficult to justify large design investments until you get production numbers high. - This is the eternal hen-and-egg-problem that manufacturers face. You have to control the risk of technology investment. This leads to auxiliary care systems that simplify care with small steps. Humanoid robot that can do the same as a nurse is quite impossible both technically and economically.

Question: Tesla revolutionized the electric car market though not a car company. Same also for google that brought us to the times of the driverless cars while – again – not a car company. Do you see some potential for a parallel development in the robot market?

Hannu: Tesla has several good innovations for electric cars. They also had courage – or stupidity – enough to let people beta test autonomous driving in real life. In addition, Tesla had to wrap the costly batteries into luxury cars, because nobody wants a normal car with batteries doubling the price of the car. Nevertheless, Tesla seems to be close to succeed. This may change, when traditional car manufacturers will be able to mass manufacture also all the new electrical components for electric cars.
The companies with fully functional robot automation – e.g. ABB and Sarcos, may succeed in making low-cost versions of their complex systems. Simple exoskeletons – like the one of Honda - may be the first mass-produced equipment with real potential in extending the time we are able to live in our own homes. Delivering exoskeletons to help in hard labour e.g. in construction, will keep the production numbers high.
Warehouse logistics by e.g. Locus Robotics will easily find direct applications in hospitals and care dormitories.

Question: At the end, and independently on how much we talk or care about the value, everything has to come down to the level of costs and prices. Given that the price of having on a 24 * 7 basis so 3 shifts minimum human carer to care for you in the distant future, or have a same priced robot, what would you prefer? No  need  to say that the robot might speak perfect Finnish that as we understand is not an easy language, while might be capable to endless play chess or backgammon with you, and also answer the very same questions many times per day, something that humans might not be able to do as good.

Marketta: This is a tricky question. I would like to have both human and robot care   The robot could take care of some routine and physically heavy tasks and also the more private tasks, and be available all the time. On the other hand, I would like that a human carer also comes to see me every now and then to check that everything is right. Humans are good at social intelligence and creativity and so better problem-solvers when I have changing demands.

Hannu: I prefer robots, because I worked designing them – and their software - from 1983 to 2010. I have been mostly employed by the state of Finland. Thus have not collected a fortune needed for three shift human carer. Robots – and intelligent speakers – will soon speak Finnish. Note, that Google just released a phone with 40 languages “on-line” spoken interpretation. Finnish IS included. I would not play with computers, but would certainly be more interested in hearing the latest news and discuss politics for example. You may have noticed that ability to speak is one of our most preserved capabilities during aging.

Question: In MARIO project, our robots are supposed to fight loneliness - how would you define loneliness? What is loneliness for you? And how would you think a robot helps to fight it?

Marketta: Loneliness to me is that a person does not feel connection to other humans and believes that she or he is not meaningful to others. I think a robot can help to alleviate loneliness a lot but not totally to overcome it. A robot could facilitate, encourage and motivate the person to get in contact with other humans, or even help the person to learn social skills. But I am not sure if a robot can provide a real deep connection or relationship, for instance because we would know it is programmed to do so. So the robot friendship would not feel genuine in a human-like way. But even if robots could not be human friends, they could be robot friends, different and better than we can now imagine.  

Hannu: At the end, we need a companion to chat with. My aim is to live so, that I do not need to be bitter at the end. I will do everything as long as I can … and then relax. Then I will let the robots do the practical things e.g. to connect with my boys and to fix coffee – if I cannot do that myself. From my point of view, we have about 20 years to fix those robots.

I hope that my contribution is sufficient … for fixing them.

 

Hannu LehtinenHannu Lehtinen studied electrical and control engineering and robotics for his Master’s Thesis on robotic welding in 1983 at the Helsinki Univ. of Technology (now the Aalto Univ.) He visited ASEA Robotics AB in Sweden for 3 years from 1985 programming robot operative systems but otherwise developed robot and automation applications at VTT from 1983 to 2010. Principal topics were industrial robot applications, walking machine and autonomous vehicles. The topic of his  Doctor’s Thesis in 1994 was force control of a walking machine. Lehtinen has been a technology, patent and market analyst at VTT from 2010. Lehtinen has also been a co-owner of  a Michelin rated restaurant for 6 years.

 
 
Marketta NiemelaDr. Marketta Niemelä is a Senior Scientist at VTT Technical Research Centre of Finland. Currently at VTT she is leading the research topic of Human-Robot Interaction for service, social and care purposes, consisting of three large-scale national/international collaborative projects. Her particular interest is in the acceptability of robots in society. She is educated in psychology (Ps.M.) and completed her doctoral dissertation (2003) on  Human-Computer Interaction in information systems and computer science.