Sitting in a studio in Kensington, London, the designer Sebastian Conran walks me through a worst-case scenario. “Basically, what it’s looking for is a break in routine,” he explains, pointing to a drawing of an elderly woman, collapsed on the floor of her home. “There’s an event. The e-sensor in the room notices that you’ve fallen over. MiRo goes to investigate.”
MiRo is a robotic dog. There is an early model close to where we are sitting. Its head sits above a torso without arms or legs, and its cartoonish eyes stare out below alert ears embedded with speakers. Conran’s company describes it as a biomimetic companion robot, and says it will eventually work with facial recognition technology to make life easier for its owner – to prompt them to take medicine, or to remind them of visitors’ names, or to question them if it thinks they’re in trouble. Conran tells me to think of it as a cross between a pet and Radio 4’s John Humphrys.
“It tries talking to you,” he continues, “and then it will send a signal to the hub saying there seems to be a problem. The hub will then broadcast on the home speaker, asking again if you’re all right, and telling you to slap your wrist.”
Conran points at the wristband the woman in the picture is wearing. It is another key part of his system, monitoring the vital signs of its wearer 24 hours a day. “If you slap your wrist the process will stop but will be logged,” he says. “If you don’t slap your wrist it will break through security, and go to a carer, who can see your heart rate and body temperature, and rewind your life using the cameras in the home to see what happened. So when the ambulance gets there, they’ll know what they’re working with.”
In this scenario all aspects of the house come together – the sensors in the room, the wristband, the biomimetic companion robot, the data hub, the cameras embedded in the ceiling… At other times, however, the system coalesces less firmly around the person living inside the home. Elsewhere in Conran’s studio is a robot table being developed with Sheffield University. The table adjusts itself to suit the user’s size and moves on four wheels across the floor directed by voice commands and tracking optics in the ceiling. The eventual aim is to attach a robotic arm to it, so that the table can move around the home, picking things up from shelves and bringing them to the user. Order a dinner from the apartment canteen, and the table will scoot through the building to deliver it.
“The thing about human care is, unless you’re particularly disabled you don’t need someone there all the time,” Conran says. “But you do need to deal with loneliness and fear. If a carer comes to see you twice a week for an hour, what’s happening in between? What’s happening to people when a carer isn’t there?”
Conran is not the only one to ask these questions and look for answers in the use of robots and smart systems. In Japan, for example, where 26% of the population is aged 65 years or older, Dr Toshiharu Mukai of Meijo University has been developing a robot capable of lifting patients out of beds and into wheelchairs. It’s called Robear, is designed to look like an approachable white bear, and could, if successful, do a job that would normally involve several care workers. Closer to home, in Grantham, a prototype care robot was recently introduced to a sheltered housing centre. Part of an EU-funded project dedicated to investigating the ways technology can improve the day-to-day lives of elderly people, the robot is designed in part to – much like Conran’s MiRo – combat loneliness, and to offer reminders for tasks such as taking medicine.
Many of these visions are nascent, but at a time of rising elderly populations and dwindling public resources, they raise important questions about where we are going with care. Can robots serve as a useful supplement to human contact, or do economic realities mean technology is pushing us towards a future where some of society’s most vulnerable are passed into the loving arms of machines?
Travel up the A1 from London to a leafy parish in the home counties, and you’ll find a house owned by the University of Hertfordshire. The two-storey building looks from the outside just like any other suburban home in Hatfield. There is a doorbell beside the front door, there are sofas, chairs, beds. There is food in the kitchen, there are pictures in frames. It isn’t until you glimpse a monolithic, 200kg care robot that you can tell this isn’t exactly a normal household.
“We first did studies on campus. As researchers, that’s what you usually do – lab experiments,” says Kerstin Dautenhahn, professor of artificial intelligence at the University of Hertfordshire. “What we found is that participants in the experiment didn’t really buy into the idea. What we realised is that we need to provide a more naturalistic environment. Something real.”
Like Conran’s vision of robotic companions and autonomous tables, Dautenhahn’s home is angled towards a future where machines play a large part in how we care for the elderly. She says her coterie of robots provide a number of functions, from physical and cognitive assistance to social help. Some of the technology in the home is designed to help users fetch objects, some is made to remind them to take their medicine, some of it is designed to remind a person exactly who is standing beside them.
“It’s also equipped as a smart home,” she says. “We have more than 60 sensors in that house, not only omnidirectional cameras that can identify and track people, but also various sensors, such as ones for temperature, attached to all the hot and cold water taps, so we know when someone is in the bathroom and using the taps or flushing the toilet.
“We have sensors attached to all the electrical devices, such as the TV, the kettle, the toaster and the fridge, so again we can detect when someone opens the fridge door, or puts on the kettle,” she adds. “We’ve also developed some algorithms for activity recognition, so based on the patterns of – for example – when someone makes a typical dinner or breakfast.”
Dautenhahn explains that the home observes a person’s activities, and then makes judgments based on those observations. If the house can tell that a person is making breakfast, it could send a robot to approach and offer help. The sensors, with their ability to track activities and log this in a central data hub, allow the system to make predictions about a person’s routine, and to notice if the routine has changed – if they are eating or exercising less, for example, or sleeping more. Sit in front of the TV for too long, and a robot could come around to encourage you to move.
The last thing the researchers want is for people to feel bossed around, though Dautenhahn admits it’s a difficult balance to strike. If you have a system that is created to protect an elderly person, and it notices that the homeowner hasn’t drunk a glass of water for six hours, should it say something? Should it tell a neighbour?
“At the moment I have a PhD student who is specifically investigating our sense of control,” says Dautenhahn. “For me, as a researcher, this issue has been part of our whole research agenda all the way. If you have a system in your house that is with you 24/7, monitoring everything that you do, then what you really don’t want is to have the feeling that there’s this super robot in the house – that you’re no longer in control. If you have someone all the time criticising you, telling you that you haven’t had anything to drink for five hours, you’ll ask yourself: whose house is this?”
So whose house is it?
When you consider the possibility of homes hosting a small army of integrated sensors and robots that may or may not be modelled on large white bears, you have to ask yourself whether these solutions conflict with the layout of, say, Edwardian terraced houses or postwar tower blocks. Dautenhahn bought a suburban home to test robots in a real-life living environment, but even she acknowledges that the clutter of everyday life isn’t best suited to robots doing their job.
“Few people only have laminate flooring, no carpet or rugs, where everything is nice and clean and smooth,” she says. “There’s usually lots of things people have in their environment that make it more complex for robots. This is why some people argue that it might be better in the future for newly built homes to accommodate the requirements of technological solutions.”
With this comes a big question: should the homes of the future be designed around robots? A demo in 2015 for Toyota’s human support robot – again designed to address the growing demand for long-term elderly care in Japan – shows the one-armed robot picking up objects in a bookcase lined with plastic boxes, each emblazoned with a label that the robot’s vision algorithms were able to detect. Such systems will presumably become less clunky over time, but image living in a house where you have to put everything you own into plastic boxes with labels for a robot to recognise. What will that do to a person’s sense of control, not to mention personal identity?
There is clearly a balance to be struck, but if we’re thinking about radical changes to the architecture of our homes, what other sorts of structures we should be looking at? One nursing home in the Netherlands, for example, offers rent-free apartments to students in exchange for 30 hours of work each month with their elderly neighbours. Elsewhere in the Netherlands, the village of Hogewey, on the outskirts of Amsterdam, has been built from the ground up to accommodate dementia patients. The 23 buildings in Hogewey are designed to be exact replicas of homes from the 1950s and the 1970s, each built to fit a person’s perception of reality.
Both of these, in very different ways, show that there is scope for social architecture that reconfigures the position of the elderly in society without technology. Instead of designing the architecture of our future around smart systems, should we be looking at structures that encourage cross-generational interaction, or that shape themselves in response to diseases such as Alzheimer’s? Perhaps, but there are, as always, economic issues.
“There are fantastic examples of supportive, dignified care, and examples of carefully building community links between generations that function as a support system with benefits for both sides,” says Lydia Nicholas, a digital anthropologist who has worked with the likes of Nesta, the Science Museum and the BBC. “Unfortunately right now these are in conflict with enormous pressure to immediately cut costs.
“These dementia villages employ a lot of people; they are expensive to run and have very long waiting lists. For a system under pressure it can be tempting instead to turn to technology solutions where potentially one nurse could look after many people because she has so much data on how all of them are moving and sleeping, whether they’re eating on schedule, etc. In the short term sensors are cheap and people are expensive. But people can provide that essential human touch and connection that supports our health over the long term.”
Nicholas tells me that it is unwise to think about future care for the elderly in terms of mutually exclusive visions. The aged population is not a monolithic entity but a varied range of age groups with different needs, different levels of ability and independence. It’s not a question about all-in with technology or all-in with new social architectures, but rather a discussion about how these things can work together. That said, there are concerns about how machines can work with the current human-based care system, especially when you’re dealing with councils facing increasingly precarious balance sheets.
“As costs rise we could end up with the system either spending more and more on getting the same level of human care, or ending up with a lot less, and having to work with that in different ways,” says Nicholas. “We need to be ambitious, creative and careful in these new systems.”
The technologists are keen to impress that their creations are designed to work alongside human care, not to replace it. And yet they also admit that they don’t have a great deal of say in how their technology will actually be implemented. There’s a sense that, in spite of good intentions, these techno-utopian visions could find themselves undermined by economic obstacles, not to mention fissures in our social fabric.
“There is a kind of crisis,” says Nicholas. “The question is whether you can build systems that enable people to build their own support networks. Using technology and system change as much as possible, are there ways you can connect people up to their local community? Of course health and social care challenges don’t exist in a vacuum; these issues must be addressed in the wider context of urgently needed work on community building and social cohesion in post-Brexit Britain.”
I ask Dautenhahn, as someone who designs robots that may one day be commonplace in care homes, whether there are concerns about how her machines may be used. “I completely see that there is a danger for formal carers,” she says. “The NHS for example could at some point decide the robots are doing a really good job, so maybe they can cut costs. Rather than having two visits a day, we could have just two visits a week, because, as you say, they have all the data.
“This is certainly not something I’d advocate,” she adds. “I’d personally be afraid if systems are being used to replace human contact, because there is a lot of misunderstanding about what robots are – that robots could have human-like qualities and give emotional support. For me, this is absolutely out of the question.”
Dautenhahn says she is against the idea of robots that resemble humans precisely for this reason. She tells me that there are things humans are good at, and there are things machines are good at. Posing a robot as a human could dangerously confuse these lines, potentially misleading authorities facing a strong drive to cut costs and automate processes. “I can see this danger,” she says. “Therefore it is very important that we have this public discussion.”
Back in Conran’s studio, he runs me through his vision of a smart home for an independent elderly person. He shows me a sketch of a single-floor flat. Sensors are dotted throughout, an autonomous table tends to the kitchen and a pair of MiRos wait attentively, close to the home’s occupants. If one of them should fall, the robot dog will be on hand to ask them if they are OK. The sensor on their wrist will confirm if they aren’t.
Putting aside the question of whether the potential success of such a system would inadvertently dissuade authorities from investing in human carers, is a home like this something a person would actually want? Conran mentions an interesting point. For the generations now growing up that are accustomed to digital culture, centred on their whims and aspirations, the idea of 24-hour attention may even be desired.
“These people are going to have completely different expectations,” he says. “We expect the state to provide for us in a way that previous generations maybe didn’t. Life has constantly got better and better, and we’re coming up against the fact that, instead of getting better and better, things are dropping off. Ageing is not to be recommended.”
