An ageing population brings with it the concern of how to care for the elderly. With care workers finding the strain of understaffed care homes hard to cope with, it could be that robotic companions are capable of taking some of the pressure off.
Matlda, a miniature robot that resembles the BB-8 of Star Wars fame (or a small, rounded waste bin with eyes, for the unacquainted), has facial recognition and is able to correspond with those around it. It learns to recognise people and can memorise their likes and dislikes.
Dr Rajiv Khosla, at Australia’s La Trobe University’s Research Centre for Computers, Communication and Special Innovation is part of the team that created the bot, now sold under HC Innovation.
He says: “We conducted our first trial in 2010 in Northern Queensland in a nursing home. Matlda engaged in delivering sensory services and social services to the residents – they include people with dementia, Parkinson’s, hearing and visual impairment”.
“Matlda is one of the first social support services – she can play bingo. That was the ice-breaker in terms of allowing Matlda to engage with the older residents who had never used the technology before.”
Khosla shares an example of how the robot can help. “Mary has dementia, she lacks self-confidence. Matlda will say ‘good morning Mary, would you like me to play some music for you?’. Matlda can recognise your face or eyes, and can play Frank Sinatra. It stimulates Mary and she starts to sing the same song with Matlda,” says Khosla, who adds that this kind of two-way conversation can be used to precede care given by nursing staff, or in his words “patients become more cooperative with the carers”.
Aside from singing, reading books and playing quizzes, the robot is able to detect a certain amount of movement, so that if a patient falls over, a carer can be alerted. Likewise, the robot is able to link up to heart rate sensors and other health monitors.
Closer to earth is the Q-Bot, which looks a little like a remote control car, though more autonomous: it finds its way beneath the floorboards of old homes, leaving insulation in its wake. It is the product of a conversation between Professor Peter Childs of Imperial College London and architect friend Tom Lipinski.
Childs says: “He was saying he was interested in sustainability and making sure the buildings we have are as good as they could be. The challenge with thermal insulation is getting in to confined spaces”.
The result of that conversation is a shoebox-size robot that feeds in to the cavity underneath a house. “The robot trundles around under the house and sprays insulation”, says Childs. At the moment, the company works with partners across the UK.
While they come in all shapes and sizes, some robots collaborate in the sense that they work as aides to human counterparts. Danish company Universal Robots is one of the best-known collaborative robotics companies, producing three human-like arms that can be placed in production processes for a variety of jobs such as polishing or picking up objects.
“It can work as an assistant – putting screws in, applying a sealant,” says Mark Gray, Universal Robot’s sales manager for the UK and Ireland. As he discusses the uses of the robot arms, he has just finished assessing the utility of the robot for applying icing to gingerbread men in a bakery.
“There’s a real mix of things that are possible – food, automotive, plastics, packaging – all different industries can benefit,” he says.
The robots are designed to last for around four and a half years, with the assumption that they can work for 24 hours a day. Once a trial has been conducted to assess the possibility of using them, they can be up and running in an hour and training courses are available online to learn how to use them.
“We expect them to require no maintenance,” says Gray, adding that the starting price is £17,000.
Dr Matthew Howard is lecturer in informatics at King’s College London and has a good grasp of the latest research. One hot topic is a robot that learns by example in much the same way that humans are taught. In a practical experiment, researchers show the robot that they want tomatoes picked up and placed in a pot. The robot then copies them.
“That’s something we can do right now quite robustly… [and] something in the next couple of years we’ll increasingly see. There are two parts: what is technologically feasible and where there is the demand. In horticulture and agriculture there is a major demand – for a few years [these industries have been] struggling to fill posts in low-paid work.”
Rich Walker, managing director at Shadow Robot, agrees, saying that soft-fruit picking is a current ambition of several companies. But can robotic fruit-pickers adapt to conventional farms?
He says that the reason for few commercial options at the moment is “not because robotic technology isn’t there – but other parts of system aren’t”.
One of the reasons that robotic adoption is picking up in high-value manufacturing is because it makes more sense to replace or assist higher-wage humans than low-wage ones, at the moment.
While sci-fi stories are full of intelligent robots not quite understanding the nuances of human life, it is a problem that currently underpins much robotics research.
Walker describes an engineer’s approach to a farm: “The first time you bring an engineer on to a farm they say: ‘you need to remove all this variability’. But the plants will always be different”.
He says this is a challenge for robotics, because the machines need to assess various aspects of a plant to see if it is suitable for picking. Nevertheless, he believes that these challenges will be surmounted over the next few years.