Artificial intelligence is a key component of the future of healthcare. Indeed, the era of the AI doctor seems to be unavoidable.

Today, we see AI in hospitals helping clinicians identify medical risks; predict when to provide targeted, life-saving interventions; form treatment plans for patients with rare diseases; and deliver precision medicine.

However, one of the benefits associated with AI may actually be a disadvantage in healthcare. While AI excels at making unbiased, purely logical decisions, it cannot yet appreciate the complex blend of emotional, social, cultural and physical needs of people. If left in the hands of a purely AI-driven doctor, the diagnosis and treatment plan may be correct, but it may not be what the patient will respond best to.

Human doctors understand how to treat people, rather than just diseases. They will read the non-verbal clues, understand when to probe more deeply for a better reading of the patient’s condition, and the type of care plan that is most likely to work for that individual. The human doctor builds a relationship with their patient, even if only fleeting, based on emotional connections.

If the human touch is fundamental to healthcare, can AI ever step in to play a role in the softer side of care?

Crunch time

With the demands placed on the healthcare system by an ageing population, the focus has understandably been on continued efficiency and higher productivity. However, this tends to leave less time for the equally critical, softer, caring side of healthcare.

Discharged hospital patients going home to self-care, with only occasional follow-up medical appointments, may struggle to maintain their care plans. They may be confused about what to do or how to do it; they may feel better and hence decide to discontinue their medicines. They may even forget or lose motivation. This is particularly true for seniors living alone. When you add social isolation into the mix, and the increased likelihood of depression, keeping people well at home can become a real challenge.

Despite the fact that this softer side of care is less logical and more about relationship building and behavioural nudges, it appears that AI can also play a role here.

The social robot

At the simplest level, AI chatbots at home can help patients keep on top of care plans. AI applications that remind seniors when to take medication, about doctors’ appointments, and even when to eat, can help in removing the anxiety and confusion that many of them face.

Above and beyond this, AI-powered social robots can provide some level of companionship for lonely seniors. We have seen in our own projects that seniors begin to create some form of relationship with AI chatbots. Social robots can also tailor their engagement to begin to nudge self-caring behaviours.

For example, letting seniors know about social activities in their neighbourhood may encourage them to go out of the home and interact with others, reducing social isolation. Sharing stories with health-promoting messages can entertain and subliminally provide education about healthier living. AI suggesting healthier food choices and encouraging them to get active can help to build healthier habits.

These are just some of the ways in which social robots can engage with their senior, helping the senior to build a relationship with their robot as well as sending self-care nudges.

Combating isolation

Integrating social robots with sensors worn on the body can support seniors in managing illnesses such that they feel reassured and more confident about living alone.


Integrating social robots with sensors can help seniors feel reassured and more confident about living alone.

One of the biggest fears seniors express when they live alone is that they will fall over in their home and no-one will realise. AI devices that can predict and prevent falls are evolving . This may give seniors with conditions such as Parkinson’s Disease the confidence to leave the house, rather than have the fear of falling keep them prisoner at home.

Sensors within the home can additionally determine when someone has fallen, or has encountered other accidents at home, and can notify emergency services for help, again providing reassurance and building confidence in living well at home.

The benefits of AI for this softer side of care may ultimately produce measurable impacts in our clinical systems. If AI can help seniors to care better for themselves, and to live at home for longer, resources in our healthcare facilities can be freed up.

The power of AI for health appears to lie in a combination of applications. Clinical applications will continue to be invested in, and will become increasingly powerful at keeping people alive. Empathy-based AI applications to support the softer side of care could help us to keep people from requiring clinical treatment altogether.

The business case for AI in the continuum of care could be very powerful indeed, and the impact on health outcomes could be transformative.

Tamsin Greulich-Smith is Chief of the Smart Health Leadership Centre at the National University of Singapore’s Institute of Systems Science.

Main image by Nithi AnandCC BY 2.0
Image of Paro, the robot seal, by AIST