How might robots support our ageing population at home?
We speak to Dr Evangelia Chrysikou and Dr Vivi Antonopoulou about how smart technologies can enhance independence and dignity in later life.
Is our world designed to support our ageing population?
According to the Centre for Ageing Better, in the last 40 years, the number of people over the age of 65 has increased by 3.5 million in England and that figure is expected to rise.
With age comes an increased risk of frailty, an indicator of a person’s overall resilience and how this relates to their chance to recover quickly following health problems. Frailty has also been identified as a risk factor for dementia. As life expectancy increases, the number of people living with dementia is also projected to rise sharply; from 982,000 in 2024 to an estimated 1.4 million by 2040.
It has therefore never been more important for us to consider the way our homes are designed and how technology can be harnessed to better support us as we age.
Dr Evangelia Chrysikou, Associate Professor at the UCL Bartlett Real Estate Institute and project lead, explains: “The problem of housing is incredibly complex. Not only are we having to consider the fact that much of our global population is ageing but government policies and political decisions across the globe often restrict the extent to which we can adapt or modernise the built environment in response to climate change. Since much of our housing has already been built, it is not like we can demolish it all to start again.
“This means that we are stuck with the old buildings which were designed for a totally different demographic. We need to start to think about other solutions beyond building design.”
The design of age-friendly homes benefits people living with frailty and older adults yet with such limitations on building appropriate homes, technology could be another way to support us as we age.
Advances in smart home technologies, including robots, could one day help older adults to live independently but we are still in the early stages of development. Whilst these robots work well in controlled research labs, they can struggle to adapt to real-world homes, which are often smaller, more cluttered, and full of obstacles.
To find a solution to these challenges, a project funded by UCL Grand Challenges brought together experts from robotics, healthcare, and architecture at UCL and beyond. The team visited various homes, including sheltered housing for people with frailty, to gather data about real living environments. They created 3D scans, detailed floor plans, and architectural models of these homes. Using this information, they tested how robots interact with realistic residential settings, identifying challenges and opportunities for improvement.
Dr Vivi Antonopoulou, a Senior Research Fellow based at the UCL Centre for Behaviour Change, explains: “AI and robotics have huge potential for supporting people as they age, however, we often find that they fail to meet the needs of the people that they’re designed for. It’s vital that behaviour science is included in this sort of project because we want to understand how people with frailty or dementia, who have complex physical needs, can live independently and with greater dignity when assisted by robots. We want to see what does and doesn’t work for this group of people.”
With ethics approval and community partnerships already in place, the team deployed humanoid robots in two homes for one-day trials. These deployments generated observational and interaction data to refine an existing human-computer interaction (HCI) framework, captured built environment factors, robot design considerations, and the social-psychological dynamics of frailty.
Dr Antonopoulou explains: “We are interested in understanding how people respond to living with a robot in their home, not just whether it “works,” but how they emotionally, socially, and behaviourally engage with it. Can it genuinely support autonomy? Does it improve routines, wellbeing, or feelings of safety? From a behaviour change perspective, we also want to understand the role of habit formation—do interactions with the robot become part of someone’s daily life, or do they feel artificial or even disruptive?
“Building on these insights, we can map barriers and facilitators to appropriate behaviour change techniques (e.g. prompts/cues, action planning, feedback, social support) and embed them into human-robot interaction design and implementation plans for scaled roll-out, including extension to related conditions such as early dementia.”
Whilst still in the early stages of analysing the data, preliminary findings indicate a mix of positives and areas to improve. For example, the robot’s rounded, ‘sci-fi’ design enhanced participant’s engagement, though it also inflated their expectations of just how much the robot was capable of. This highlighted the need for clear capability-setting.
“The participants enjoyed conversational interaction, and saw clear value in reminders for drinking, eating, and taking medication,” Dr Antonopoulou says, “The robot also fit well in small flats and navigated without getting in the way of the residents. Key barriers were connectivity (some homes lack reliable internet), robot’s lack of dexterity (i.e. inability to handle objects like door handles) and voice turn-taking with slower, story-like speech, where users were occasionally cut off mid-sentence. Finally, there were concerns about over-reliance given the importance of independence.”
Dr Chrysikou adds: “Overall, an approachable design, voice interaction, and reminders show promise, but success will depend on reliable connectivity, more forgiving voice handling and personalisation, and autonomy-supportive features and onboarding.”
So is there potential for robots to assist people living with dementia in the future?
Dr Antonopoulou says: “We want to take this study further and design robots specifically for people living with dementia. That may include changes to the robot’s communication style or its ability to provide memory prompts for people with dementia, to recognise distress and respond appropriately by sending signals for help. Once we reflect on how our pilot study went, we will be able to look further into this.”
“As technology advances, it opens up more possibility,” adds Dr Chrysikou, “We can start to be more optimistic about the capacity of robots to one day help more people live independently in their own home for much longer.”