In the second part of the interview with Ms Tamsin Greulich-Smith, we learn about how the Centre applies the principle of human-centric design. The Centre is working on an integrated care pilot project in collaboration with Lions Befrienders, a voluntary welfare organisation, which provides friendship and care for seniors to age in place, as well as Yishun Community Hospital and Allied World Healthcare.
The project aims to transition senior citizens back into the community more quickly after a hospital stay and in a way that allows them to look after themselves better when they are discharged, through integrating them with community services.
The question being explored in the project is: Can a personal artificial intelligence (AI) device encourage seniors to look after themselves better.
The elderly can get socially isolated when they live alone, and then they may stop caring for themselves.
Ms Greulich-Smith explained, “Some seniors can start to give up, when they live alone. They don’t bother to go out because they may have lost confidence or lost purpose. Then they may get a little depressed, which can result in them losingtheir appetite and stop eating so well. Their health suffers. They might stop taking their medication or get confused. They get anxious about various things, which can affect the quality of their health.”
The AI device can help remind the senior citizens about different things, designed to encourage them to take better care of themselves. It can prompt them to eat, reminding them to start the day with breakfast, for example. It can prompt them to exercise or complete their rehabilitation exercises. It also reminds them to take their medication and encourages them to join daily activities for seniors in their local Senior Activity Centres. The system also integrates with Lions Befrienders Silver Bow campaign, which sends seniors messages from members of the community. Ultimately, through this range of engagement, the seniors might form a relationship with the technology, if it is designed correctly.
This pilot is a rolling programme of onboarding patients, based on when the patients are ready to be discharged from the hospital. The patients who seem viable to take part in the pilot are asked if they are interested. If the answer is in the affirmative, then Lions Befrienders step in to explain the project to them and assess them for their suitability. The patients are assessed over a 4-week period while they are in the hospital. Once they are discharged, and the consents completed, the technology is placed in their home which comprises a dashboard providing a shared patient view and the aforementioned AI device for some of them.
The human-centric approach in practice
In this project, the Centre has tried to adopt the human-centric approach advocated by Ms Greulich-Smith. And the results have been illuminating.
The Centre held a design workshop for seniors, in their 80s and 90s, to get their inputs for an artificial intelligence (AI) device to be placed in their homes. The team had the basic device, as in an unadorned version of the device, to help the seniors understand how to interact with it. The workshop also explored various design options with the seniors, to either conceal or otherwise improve the appearance of the device.
Prior to the workshop, most of the professionals working with the seniors had assumed that the seniors would not want the ‘ugly’ unadorned technology in their house. They would want it to be camouflaged. But the seniors chose the raw technology!
They had a pragmatic attitude. Their primary concern was that if the voice-activated device is placed inside something else, it won’t hear them so well. ‘I don’t want anything to possibly prevent the technology from hearing my voice’.
The second reason for their choice was a sense of pride in having this modern technology in their home. The device would become an attention grabber, as neighbours popping in asked about the device.
“So we cannot make assumptions about people’s motivations, desires and concerns. The design process is important because it stops us making assumptions, and helps us improve the impact of our solutions” Ms Greulich-Smith said.
Embracing failure
In order to quickly prototype a model in the first stage, an off-the-shelf device was used for the AI component. Though there were concerns that the off-the-shelf device was not doing a lot of the things the seniors want or require. For instance, the device doesn’t speak or understand Hokkien or Singlish. Which requires the project team to find seniors who can speak English, and there aren’t so many of them. And if they do speak English, their accent isn’t easily understood by the device. However, the flaws in the off-the-shelf solution need not be cause for concern.
“I said, well I think that’s fantastic. This is our sacrificial concept. This means that we are creating this very long list of things we will design out in our own model during subsequent phases,” Ms Greulich-Smith said.
She explained that the key thing about prototyping is having the bravery to welcome failure and make the list of failings as long as possible. Because the longer it is the better the design will be in the next phase. The aim is to learn from the positives as well as the negatives.
In addition, the team observed that the seniors are loving having this this technology in spite of all the flaws. So, if those failings can be designed out, the team might have something that the seniors will genuinely start to have a positive relationship with.
“If we scale too soon, we have an expensive mess to fix. But if we just have a small pilot perfectly intended for prototyping and testing, then we are allowed to fail.”
Scaling up
Once the team has robust evidence that the approach is working, then they will seek funding either through their own institute or through a grant-funding body to explore scaling up. At the moment, the project is focused on one geographical area and patients from one community hospital. It could be
expanded across the island and to more community hospitals. Taking a different tack, acute care hospitals could also be considered which present a very different situation.
In the second phase, the team will build its own AI device tailored to meet the needs of Singaporean seniors living alone at home. This will be based on the learnings from phase 1.
Ms Greulich-Smith said, “Our plan is to have an opportunity to eventually, nationally scale part of that model, if not all of it.”
There might be privacy concerns and challenges once the project is scaled up. At the moment, it is explained to the seniors that the data is only used within the project and it is not shared. In the next phase, researchers might give the seniors a download of the questions they have asked and let them remove anything they don’t want others to see.
The shared patient view is used as a collaboration and engagement tool, for caregivers and next of kin to monitor their condition and check things like if they are having their meals on time. So, it is important for the seniors to understand that this information is being shared and to get their consent for who has access to the data.
With such challenges, the Centre’s approach is to explore them with a small group and spend time in understanding the issues as thoroughly as possible, and then responding to the issues that arise in an agile way.
A blossoming innovation hub
We brought up the topic of other similar projects in Singapore to help seniors age in place, some involving AI as well.
Ms Greulich-Smith reacted saying it’s wonderful. “It’s what we need. We need an island full of innovation, people experimenting and ex
ploring. That is really what is going to start to find the breakthrough solutions.”
“As soon we start trying to control that too much, we will lose the innovation culture. At the moment, Singapore is really blossoming as an innovation hub.”
She added that the Centre is always willing to share its findings and help other organisations with whatever projects they are conducting.
The Centre actively undertakes projects with different partners, and seeks opportunities to cross-fertilise across them. For the integrated care pilot, subsequent phases could see the Centre trialling adapted approaches with different partners to explore improvements and challenges in different settings. Over time, more partners are brought in to enhance the offering. The entire development process is collaboration-driven.
“Let’s see what we can learn by all of us exploring in different ways. Because there’s no one answer to the complex problems in healthcare,” she concluded.
