Source: Radio New Zealand
Pepper is a robot designed by Aldebran to specialise in communication and interaction with humans for situations including providing companionship for elderly people. RNZ / Philippa Tolley
The global population is ageing, and New Zealand is no exception. Almost 20 percent of Kiwis are projected to be 65 or older by 2028, and like many other countries around the world, Aotearoa faces a shortage of healthcare and care workers to look after the elderly.
Could robots with the ability to perform everyday personal and household tasks help meet some of those needs?
Robots have some distinct advantages, robot designer Rich Walker tells Mihingarangi Forbes, speaking from London – but there’s some challenges and hurdles to overcome, and some big ethical considerations.
As the technology evolves, do we need to start asking ourselves how far we want it to go?
Walker is director of Shadow Robot, a UK company specialising in the design and manufacture of robotic hands. He is also an industry advocate for the ethical use of robotic technology.
Walker has played with robotics since being introduced to small robots at a computer camp as a child, where he tried to use them to move chess pieces around.
“I think if you grew up with science fiction, robots are a kind of natural lure, there’s a whole idea that you could build something that could be a companion, an assistant, a helper – a fun character to have in your life,” he says. “And then you find engineering, and you discover that it’s nothing like that, but it’s a lot more interesting in some ways.”
What robots can and can’t easily accomplish are widely misunderstood, because we look at it through our assumptions they will be able to behave like a human body.
Rich Walker, with a dexterous hand. Robotic hands are a particularly difficult technology to create. Charles Gervais
In Japan, robots have been used in aged care facilities for more than a decade. But even so, he says “we’re a long way away from the dream of a robot that can wander around your house and do everything.
“I think what happened in Japan is they said ‘let’s give this a try, let’s get robots into care, let’s see what we can do and see how well it works’. And as you might expect the results are mixed. There are some places where actually you can do something quite useful, quite powerful and quite important, and then there are other places where people are just scratching their heads…”
Care work needs sensitivity, touch, judgement and gentleness, which are difficult for a robot to match our human capabilities for.
“If you look at a factory, you see the machinery in there – the robots in there do amazing things, over and over again, exactly the same.
“But if you’ve ever tried to wipe someone’s mouth, that’s never the same twice, that’s never the same experience, it’s never the same person you’re doing that with,” Walker says.
“So that’s really the problem for us, is: What are the jobs where a machine can be really useful and help and free people up to do other things? And what are the jobs where honestly you want to feel that you are engaged with a person – you are interacting with a person there?”
Some areas of human capability and some tasks are particularly difficult for robots – controlled and measured touch is at the top of the list. For example, robots still cannot use scissors.
“Partly it’s because what we do with our fingers is such an innate part of our brain, we have huge trouble thinking about it.
“We have this joke in robotics that if you ask a person in robotics how they’re holding their pen they’ll drop it straight away, because when you think about how you’re holding your pen you can’t do it anymore because it’s completely unconscious.
“And the problem with that is we said ‘oh yeah, it’ll be really difficult to get computers to play chess’ – well actually it wasn’t that difficult, you just needed big computers.
“But it turns out it’s really difficult to get robots to make a chess board, or set up a chess board, or indeed work out where the chess board is in the house and go and get it and come back, because the world is a very complicated and unpredictable place and robots work best in places that are … simple and well organised – not like my house.”
Designs to meet real needs
When parsing out the needs of aged care there have also been some big surprises, Walker says.
“We have conversations with local councils … they say things like ‘our biggest problem is how do we get someone to your house’. It’s not what they do when they’re there, it’s the travel time to get from one person to another to another.
“If you have to have three or four care visits a day, and someone has to travel half an hour for each visit, that’s quite a lot of the day taken up in that.
Things that can make a big difference can sometimes be quite simple, he says.
For instance, a washing machine isn’t normally thought of as a robot, but it is – and appropriate design can make a big difference when it comes to washing clothes for people living with incontinence. Or televisions or cell phones designed to have only a few buttons and channels are much more accessible for people with dementia.
“And a lot of this work, it’s not about clever-clever technology. It’s about saying what can we do to give this person back the independence they are starting to lose.
“And then, when they have carers come round, how can we make it so the carer can spend the time on the human element, not doing a mechanical task that could be done by something else to actually engage with the person, to give them dignity.”
Globally, the need is huge, Walker says.
- Hospitals short an average of 587 nurses every shift last year – report
- Aged care sector in crisis
“This is a big challenge, it is a big problem. Almost every country in the world has the same problem, their populations are ageing, their infrastructure could be better. Robot technology will play a big part in that – and particularly … infrastructure.
University of Auckland research tested using a robot called Bomy at two Auckland retirement villages, to help with daily routines. supplied by University of Auckland
Things like: “Self-driving cars, robots that can repair bridges or build roads or repair railways or just inspect railways. These will make a huge difference. And some of these things will free up humans so they can do more human tasks.”
Even with a belief that this is the direction things are going, the acceleration in commercial manufacturing of some robot appliances has been a surprise, he says.
“We’ve been very surprised by how cheap the recent wave of robots coming out of China are – and we’d always said that robots are going to be very expensive, that they’ll be something that governments buy and councils buy, and companies buy.
“But actually it’s starting to look like maybe when you manufacture them in very large quantities they don’t have to be super super expensive.
“So I don’t know yet how we will end up having a society where people have robots in the homes – I mean we’ve seen robot vacuum cleaners, and we’ve seen robot lawn mowers, and there are robots like that, they do exist, so it is possible.”
Caution warranted
Despite the promise, it’s sensible to take a cautious approach on robots, he says.
“If I tell you I’m putting something that weighs 150 kilos and will move at 7 miles an hour in your living room, you’re going to want to know that that’s not going to trap you up against a wall and stop, right. Because that would be very very difficult for you and possibly dangerous.
“So people who are building systems have to go in and say: ‘How do we make these systems safe, and how can we make sure that people trust them? – And you can’t build trust by saying ‘I know better’. You have to build trust by saying ‘how can I show you that this is safe and reliable and robust?’
One of the most impressive robots already being used in care environments is a small furry seal called Paro, which was designed as a companion for people who could no longer have a pet or handle one.
Paro the seal, a robot companion animal.
“It gets used in care for people with dementia. It’s a little thing, it sits on your lap, you stroke it. It’s quite warm and it has a couple of little movements it makes and it makes a thing a bit like a purr, and it’s very comforting, and it’s a robot.
“No-one’s going to feel threatened by it, no-one’s going to feel scared by it, and there’s not much it can do to go wrong, but it has fantastic value to people’s well-being… this is something they can cope with.”
Walker says any discussions about robotics for use in aged care need to include consideration that companies creating and selling robotics must earn trust – “it’s not an automatic”.
Regulation is necessary, he says.
“Particularly when we’re dealing with people who are vulnerable or in need of additional support or help, that we make sure that what they’re getting is right.
“I’m not a fan of government, but I recognise that in this case you have to start by having legislation, regulation, laws that say these things must be safe, this is how they must be safe, they musn’t be deceptive – the robot shouldn’t pretend to be something [they aren’t].
“There’s a whole package that needs working out, how we treat these things, because they will come into our lives and we need to make sure that we benefit from them.”
Is it healthy for the lonely and isolated to develop a relationship with AI and robots? It’s a fascinating question, Walker says.
“I don’t think any of us would have guessed how all-encompassing it’s possible to be with ChatGPT and with tools like that. There’s a long history of that in robotics, in artificial intelligence, where people make things you could chat to that seemed intelligent and people do get sucked in, people do really enjoy it.
“And I think there’s a wider question there – it’s like false advertising, we have laws about advertising, you’re not allowed to promise things that are not true of your products. Are we being promised things that are not true around the chatbots like ChatGPT? – I’m not sure.
“But if we are then we should definitely make sure that doesn’t happen. Because the last thing you do want is somebody who is sitting at home with a useful system that can help them, but is deceiving them, is playing mind games with them. Because it can, and it’s able to. And if we haven’t regulated that, that would be a great shame.”
Technical challenges and milestones
Shadow Robot’s Dexterous Robot Hand using a delicate grip to grasp blocks. Matt Lincoln
Some of the Robotic hands Walker works with have more than 100 sensors and have reached the milestone of being able to solve a Rubik’s Cube using a single hand.
The programming behind that is just as crucial as the physical design, and determines how it puts the physical capability to good use, he says. And after the programming there is one more step, training. That teaches the robot which of the things it is capable of are doing are things you want to happen, and details like what order to do something, or in what manner.
Robots can be trained using reinforcement learning, Walker says.
“You do the same as if you’re training a pet … you give it a reward when it does the thing right, and it gradually learns over time what things are right and what things are wrong.”
Walker is taking part in the UK government’s ARIA Robot Dexterity Programme, a high risk, high reward swing at finding ways to solve challenges in robotic dexterity, to create more capable and useful machines.
“Really, it’s the absolute cutting edge of materials science is trying to make things that behave like the muscles of the human body.”
Skin is another fascinating problem, since human skin grows back if it’s damaged, or it can thicken and become more robust by forming calluses. He hopes new learning will come out of the intersections between biology, medicine and robotics.
“Those processes through which things recreate themselves so we can continue to use them, that’s again one of those things where we go ‘wow, if we could do that, it would be transformative!’”
Sign up for Ngā Pitopito Kōrero, a daily newsletter curated by our editors and delivered straight to your inbox every weekday.
– Published by EveningReport.nz and AsiaPacificReport.nz, see: MIL OSI in partnership with Radio New Zealand
