Artificial intelligence could soon be helping carers take care of the elderly. MATTHEW SACCO, who recently completed his Master’s in Computer Science, explains his role in the project.
On average, people live longer now than they did in the past. That is both a fact and a blessing. Even so, an aging population comes with a number of issues, one of which is how we can reach a compromise between ensuring that the elderly are well looked after without taking their independence away.
This is something Matthew Sacco had to grapple with as he joined a team of individuals and entities located in Malta and Sicily to work on an EU project called NATIF Life, whose aim is to create the assisted-living apartment of the future.
“The project trickled down to me as a student doing my Master’s Degree in Computer Science,” Matthew explains. “From my end, I used a special type of camera termed as RGB-D, which has depth-sensors that can determine how far an object is from it. A system made up of a number of such cameras, coupled with Artificial Intelligence smart enough to detect what the people in a room are doing, means that we can now have an indoor localisation system that can tell when a person is in trouble or is simply going about their day-to-day business.”
Setting the camera system up in a laboratory within the Faculty of ICT, as well as at a care home in Sicily, Matthew taught the software how to tell whether a person had fallen down and required help, among other skills. Moreover, by speaking to relatives of elderly persons who live at Sir Paul Boffa Hospital, Matthew realised that the software could help relatives with some often forgotten issues.
“Many relatives mentioned that some of the hardest things to keep track of are whether the elderly in their care had been active, had eaten, had taken their pills, or done their exercises. It’s not something we had really taken into account when we set out to work on this project, but they’re pertinent issues, particularly as these all help elderly people remain strong,” he continues.
One of the biggest hurdles of the project was to ensure that the privacy of the individuals living in the assisted-living apartments of the future would be safeguarded – something which is already a major headache with CCTV and surveillance cameras.
To achieve this, the first step was to use the type of cameras they did, which do not ‘see’ colour and don’t have the ability to recognise one person from another. Moreover, while the information currently being retrieved from the laboratory here in Malta and the care home in Sicily is going into a centralised system for data collection, this would only be done with explicit consent from the elderly people in question.
Nevertheless, the potential for such a project is huge. Firstly, this could help keep track of how elderly people are doing when no one can physically be around, a particularly important factor for those families which may not afford a live-in carer. Secondly, the system can be programmed to automatically inform the relevant people and authorities should something go wrong – in case of a fall, a fire or a burglary, for example. Thirdly, when used at care homes, the system could act as a second pair of eyes to help carers with their work.
“Indeed, while the system is ready for day-to-day usage, the probability is that it will have its roll out in care homes. There it could truly benefit the elderly as many such homes are finding it hard to recruit enough personnel and it’s not always possible to be with each patient at all times. In certain situations, such as near staircases, where most accidents tend to happen, this could truly become a life-saving project.”
The system, in fact, is already being used for such purposes at the aforementioned assisted-living property in Sicily, which is a temporary retreat for the elderly who may need a half-way house while recovering, among other reasons.
The best part of the project, however, remains the way Matthew designed the system. Its modular nature means that, in the future, people could create new applications that would integrate seamlessly, thus adding to the range of features the system can support.
“There are already more research studies being proposed in order to update the system and I look forward to seeing what I and other researchers manage to do with this framework,” Matthew concludes.
That makes two of us, Matthew!