Thursday, 17 April 2014

Keeping up with the Jetsons - Long Version

Last week, the BioCentre on Ethics hosted a teleconference on the Use of Robotics in the Care of Older People. I was on a panel with Dr Walter Greenleaf from Stanford University and Dr Amanda Sharkey from the University of Sheffield.

We teamed up to discuss whether robotic assistive devices are more likely to empower independent living or intensify isolation. We barely scratched the surface. Dr Greenleaf opened the conversation reminding us that at the same time as we are facing the acute problems of caring for ageing populations, we are already living in an assisted world. He advocated for immediate action in the form of targeted, solution oriented efforts and user-led ie older person involvement in the design process. Unless actual users can be involved in all elements of design, we will be unable to stay far enough ahead of the predicted, large-scale physical and neurodegenerative inevitability's. Dr Sharkey also acknowledged the predictable expansion of robotic options, in all domains of care, but cautioned against the inherent risks, particularly with respect to end of life care.  Our overenthusiastic acceptance of introducing human-like but not actually human helpers could qualify as an infringement of personal dignity, in our older years. The subtext to our exchange of viewpoints, as academics observing these developments, was on ethical isolation amidst this burgeoning robotic innovation. Worthy of note, the concept of robo-ethics is a mere 13 years old.

The fact is, the robo-genie is way out of the bottle and robotic intervention is already occurring in every dimension of modern life. What's needed is targeted attention on the actual application of robots to provide assistance and care for older adults. Despite the robotic inventing taking place in specialized, well-funded labs and the growing awareness that the planet is ageing, a serious lack of consumer ready solutions remains. Robotics for older people isn’t core to any business plans of the relevant players in this market. This is relevant because we live in a commercially driven world.
Clearly, the business of robotic engineering and the business of caring represent two distinct cultures that need to comfortably coexist and these cultures are already intersecting.  Since 2000, amazingly complex surgical procedures using robotic arms and a surgeon operating from a console have allowed for enhanced dexterity and more refined movements than were ever humanly possible . Add to this the dual bonus of tremor reduction and faster recovery times for patients and there is not much not to like about these robots!  Next, it is well documented that most people never do any rehab exercises after they leave the hospital or physio appointment. There are now virtual avatars carrying out routine checks to see that patients are actually doing their exercises and advising them on proper technique. 
These "relational agents" are also at the forefront of medication management, "watching" patients take the right meds, at the right time. This is hugely beneficial, given that 50% of the people to whom 146M are dispensed in the US daily are never taken, at a cost of some $2.9B annually. In addition, diagnostic and pharmaceutical errors result in 700K drug emergencies every year. This translates to sunk costs to the pharmaceutical industry of £180B. Big Pharma flag bearer or not, this is material. Another recent study shows that AI, artificial intelligence algorithms, outperform trained physicians in terms of diagnostic accuracy by 41.9%. In these examples, there is proven synergy when artificial intelligence and robotic engineering are combined with humans and we leverage all the new technological tools at our disposal. The nagging question, then, is will these interactions alter the users: surgeons, carers and the cared for's basic understanding of themselves and if so, at what cost?

There is something called the Turing Test that robots can take. It's a test for intelligence which requires that a human being should be able to distinguish the machine from another human being by using the replies to the questions put to both. Many studies have shown a natural disposition of people (especially, older people and children) to accept life-like robots readily as peers, despite the fact that they are machines. This opens up the debate to concerns over inauthentic relationships.
Over ten years ago, research was conducted on an entertainment robot animal. AIBO was being used for treating severely demented people in a care home in Japan (Tamura et al., 2004).  The conclusion was that the AIBO was an effective rehabilitation tool, providing positive outcomes to patients. In addition, AIBO was observed to avoid any potential danger or injury to the patient and because it was not a real animal, cleanliness was more easily maintained. AIBO was also attributed with the most positive outcomes, when the AIBO was dressed. Apparently, the introduction of humanoid characteristics and demonstrating their usefulness in predictable environments, not situations requiring intuition or spontaneity, is the implicit driver of good tech for older persons.
Similarly, a "welfare robot", in the form of a robotic seal called Paro has also shown beneficial effects on people with dementia. A landmark Danish study aims to professionalise the use of Paro by certifying professional caretakers in their use of the robotic seal. The objective is to determine the extent to which new technology can support care home residents’ independence, communication and needs for stimulation. So does the robotic seal give care staff more time for social contacts with other residents, thereby enhance the well-being of both parties or  is this actually a clever distraction, further limiting personal contact?

And what about loneliness? In the UK, it is estimated that over 50% of over 75 year olds who report living alone, say their TV is their main company and 17% report that they only have contact with a person, less than once a week. If you are being cared for by social services, your carer may not even speak the same language as you do and can only spend, on average, about 15 minutes, per day, per visit.
This makes it hardly surprising that neuroscientists have found higher levels of epinephrine, the stress hormone, in the morning urine of  “lonely people”.  Their research also showed that “loneliness” – subjective social isolation – alters the basic process of DNA transcription. We know that gene expression also disrupts perception, behavior and physiology. Loneliness burrows deep and can become a trap that reinforces isolation. On the positive side, loneliness and possibly depression have both been observed to be relieved using AIBO or Paro and in much the same way as a communication technology such as Skype works to create connectedness in people without cognitive deficiencies.
The bottom line is this, when you introduce a robotic assistive device into the mix you need user led design. Because older people respond differently to any technology than do designers or technology saturated children, they need to be included in any R&D effort. Worthy of mention, the prototypical designer of anything from F1 to a toilet seat is a 38 -year-old-white-male. What do you suppose they would rather be working on?

Is there an answer? In the case of people with dementia or other psychopathologies, are we deceiving them? If a robot is capable of impressive caring behaviour, is that wrong? Are carers who are good actors and exhibit impressive caring behaviour any different? In the same way that the carer may die, the robot may fail. Is one disappointment bigger than the other, in the mind of the cared for? Is there a distinction? Is artificial empathy, imbued by human designers fundamentally wrong? For that matter, is first person subjective experience or at best, realistic conscious like behaviour, necessary for caring? I think not. However, human intervention in terms of monitoring will for the foreseeable future, continue to be important.
Meanwhile, in all of the popular and stunningly poignant TED talks by Henry Evans and Hugh Herr or other equally important personal technology news, we see enthusiastic people fully dedicated to reengaging with the human race. These are the fortunate survivors, all living new lives and au fait with their biomechanically engineered prosthesis and body surrogate robots. Some of these come with $400K price tags, clearly out of reach for most, but they also inspire great hope. These are  authentic but they should not be confused or compared with the radically different experience of ageing.
With ageing, function is lost unevenly and inequitably, not suddenly and over the course of a lifetime. The application of robots in care of older adults is uncharted territory. The needs of an ageing citizenry are no less immediate or urgent but are unlikely to be perceived in the same way as the otherwise disabled group. The extent to which robotic innovations will assist or replace humans is yet unknown and the long term costs of such care are also unknowable.
The ultimate goal of human-robot interaction in the care of older people is to enable communication with real users in real-world environments. We need to start training a new generation, combining specialists from both ends of the age spectrum with specialist knowledge about technology and of actual ageing. Currently, there are serious gaps in the linkages between product and the population they are intended to serve (Hudson, 2014).
At present, the boomers are watching their own ageing parents and worrying. We see the centenarians, clutching their birthday greetings from the Queen and delight  to see them being adequately celebrated, obviously cared for. There is no hint of fear, no whiff of loss of control, no suggestion of isolation, loneliness or neglect.  They also illuminate that the benefits of ageing are far from equitably distributed. This has further philosophical and spiritual implications for the broad definition of caring and even categories of the notion of “elder abuse”, be it inflicted by man or machine.
By definition, these issues are intricate, vital, and incalculable. This is as true today as it was for the 1962 Jetsons and their indefatigible robot house-maid, when Rosie the Robot got burned into our boomer memory banks. She was a reject, an outdated model from U-Rent A Maid. The Jetson's only ran for 24 episodes but in that brief time, it fired our imaginations. Despite the fact that Rosie was past her sell by date, she ran the house while simultaneously dispensing wisdom, parental advice and discipline. If having a robot like Rosie, who understands where the older person is located in the home, provide them with whatever help they want or need and provide companionship, can this be so wrong?
Ageing is not a homogeneous process, care needs are highly personal and fluctuating, while eventual decline is undeniable. This debate only serves to underscore the first world nature of this problem. Older people without cognitive decline have high demands for a high quality of life. They need help with basic functions including cleaning, washing and shopping that robots of the future will be able to perform. Integrating technology with existing health and social care services makes intuitive sense but it represents such an ambitious undertaking it is barely conceivable. Meanwhile, sophisticated assistive care is, at present, available to the few: the deep pocketed and participants in optimum conditions, in controlled settings. Until sufficient interest, improved access and affordability gaps are closed, solutions loom just out of reach (Hudson, 2014).  But this also represents a yawning ethical issue and it's racing to the fore. As such, it is rightly under exploration. 
Get a move on Rosie.

Deborah Gale

For more information, the BioCentre on Ethics held a teleconference on the Use Of Robotics and the Care of Older Persons on April 11, 2014, Deborah was a panelist. Link to the audio and full report:

HUDSON, R. B. 2014. Aging and Technology: The Promise and the Paradox. Public Policy & Aging Report, 24, 3-5.

JOHNSTON, S. 2014, Aging2.0 Flipboard. Accessed: 23 March 2014. 

TAMURA, T., YONEMITSU, S., ITOH, A., OIKAWA, D., KAWAKAMI, A., HIGASHI, Y., FUJIMOOTO, T. & NAKAJIMA, K. 2004. Is an Entertainment Robot Useful in the Care of Elderly People With Severe Dementia? The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 59, M83-M85.

No comments:

Post a Comment