There are over 850,000 people living with dementia in the UK and this is set to rise to over one million by 2025. This is putting a significant pressure on the care system, carers, and families. Agitation is common in people with dementia and can manifest verbally and non-verbally and has many factors in its cause reflecting unmet needs, environmental stimuli, psychological needs and as a reaction to others. Current treatments focus on therapy and pharmacological treatment, but in people with dementia trials of numerous drugs have been shown to not be effective or to cause harm. In older adults, coexistent frailty, other chronic health conditions and the need to take other prescription medication further augment the risk of considerable harm.

Techniques to comfort and distract are advocated prior to consideration of drug intervention and yet there is a lack of non-drug strategies with a strong evidence base and therefore tackling the behavioural and psychological symptoms of dementia and hyperactive delirium represent an urgent and unmet need. However, therapy is expensive and pharmacological interventions often have undesirable side-effects. Healthy people also regularly suffer periods of anxiety. A non-pharmacological, intuitive, home intervention is complementary to other treatments and beneficial for non-clinical groups. Existing at-home anxiety aids, such as guided meditations, typically employ visual or audio stimuli to guide the user into a calmer state. However, the tactile sense has the potential to be a more natural modality to target in an anxiety-calming device.

What did the project involve? 

This project sought to develop a soft robotic device as a therapeutic product with intelligent sensing to support people living with dementia. Previous EPSRC research had developed early-stage haptic technology that offers an accessible alternative for easing the state of anxiety. This project sought to accelerate the impact of this development in haptic technology for anxiety relief and develop a proof-of-concept soft robotic health and wellbeing device that can last a user’s lifetime by adapting to the user’s changing needs and preferences.

The project brought together a strong interdisciplinary team with academic experts in artificial intelligence, soft robotics, sociology, medicine, design and healthcare, and partners with expertise in user experience and e-textiles, and art and illustration. Drawing upon the interdisciplinary team’s expertise, this project specifically sought to test the hypothesis that ‘failure to use’ results from misalignment between the profiles of users (lived experiences, needs and expectations) as perceived by the roboticist(s) (e.g., engineers and designers), and what robots can presently do as perceived by potential users. Collaborating in part with Bonnie Binary – a local textile business who “create sensory huggables to comfort, delight and enhance wellbeing”, the team conducted a focus group with people living with dementia and their care partners and a foresight workshop with roboticists with expertise in ‘soft robots’, in order to:

1. Understand how people living with dementia and their care partners think soft robots might improve their lives.
2. Explore new ways of bringing the fullness of lived experience into earlier stages of (soft) robot design.
3. Assess how this information produces actionable information for the engineers and designers and why (or why not).
4. Expand and deepen the exchange of social scientific and technical knowledge about possible trajectories for robot design in health-social care by developing a novel taxonomy of gesture in human-robot interactions (HRI).

They project’s overarching aim was to explore how we can use intelligent sensing to personalise a health and wellbeing soft robotic device such that this device responds to its user, adapt to their needs and preferences, and yields a satisfactory experience.

Specifically, the project investigated the following questions:

1. What does a soft robotic wellbeing device need to sense and understand about its human user (a) to personalise its interactions with the user; and (b) to respond in a sensitive and intuitive manner such that the user feels comforted and cared for?
2. How can we design a soft robotic device that can last the lifetime of its human user in that the device learns to adapt to its user’s needs as they change?

The project was split into three main stages:

Stage 1: Focus Group

In stage 1, the team conducted a Focus Group of 6 – 8 people with different dementia phenotypes and their carers in order to create a figurative ‘blueprint’ for a soft robot. Analysis of this then considered different uses of gesture in interaction with the material, and its scope to contribute towards a taxonomy of gestures and verbal terminology according to soft robots and/or materials.

Stage 2: Foresight Workshop

In stage 2, the team convened a half-day workshop. The workshop brought the team up to date on current soft robotics research to explore future intelligent sensing and interaction ideas. Secondly, the workshop took the Focus Groups outputs forward. Analysis of this concentrated on how knowledge, needs, values, and technological limitations were to be negotiated in these decision-making processes.

Stage 3: Sensor development for soft robots and agent-based simulation testbed

Equipped with the information and learning from Stage 1 and Stage 2, the team developed a soft robotic prototype with sensors including pressure, temperature, vibration, and tilt. They conducted a small-scale study with users interacting with the prototype in a laboratory setting. The prototype collected and generated a stream of user’s internal state and activity data as the user interacted with the prototype.

The team developed an agent simulation testbed that enabled product designers to evaluate the soft robotic prototype under varying user contexts, changing needs and preferences. From the understanding and insights gained from Stages 1 and 2, the team created agent models for human user and soft robotic prototype and seed them with data from real users obtained from laboratory experiments.

Who are the team and what do they bring?

• Nirav Ajmeri (Computer Science, University of Bristol) is a researcher centred on Artificial intelligence, their research interests are AI, intelligent agents, and multiagent systems with an emphasis on ethics, cybersecurity, and privacy.
• Jonathan Rossiter (Engineering Mathematics, University of Bristol) is a researcher and the head of the Soft Robotics group at Bristol Robotics Laboratory. Their research also include bio-mimetics, artificial intelligence, composites, sensors and the wider robotics field.
• Peter Winter (Sociology, Politics and International Studies, University of Bristol) is a sociologist in the field of Science and Technology Studies (STS) specialising in the analysis of complex sociotechnical systems, particularly sociotechnical systems involving Artificial Intelligence (AI) applications.
• Emily Henderson (Bristol Medical School, University of Bristol) is a researcher focused on Ageing and Movement Disorders. She leads the Ageing and Movement Research Group (AMRG) in Population Health Sciences running a translational programme of research in neurodegenerative disease, falls and cognition.
• Rebecka Fleetwood-Smith (History, University of Bristol) is a researcher with a focus on the sensory past, present, and future of NHS hospitals in England. Her background is in fashion textile design and psychology, particularly exploring the significance of clothing and textiles to people with dementia, using sensory and creative research methods.
• Annie Lywood (Bonnie Binary) is a designer and founder of Bonnie Binary that has pioneered research into E-textiles, exploring the potential of soft, interactive, and decorative textile interfaces that enable older people to access and manage their immediate environment and memories.
• Sally Stevens is an animator and illustrator based in Bristol. She works with paper cut-outs to make animations about the world we live in, to encourage views from different perspectives. Sally Stevens’ Website.

What were the results?

At successful conclusion of the project, this impact accelerator award delivered a proof-of-concept soft robotic device that the team could leverage for larger scale trials. They will apply for further funding opportunities to explore how we can use intelligent sensing to create responsive intuitive soft robotic products for wellbeing and the healthcare sector.

With learning from the workshops and interactions in the inter-disciplinary team, the team learned abstractions for inclusivity by design. They will seek funds to develop a novel research methodology to support inclusivity by design around these abstractions where we take best practices from social science research to support inclusion (by design) of people with various impairments when developing assistive soft robotics devices.