The population of Scotland is rapidly aging and, within the group of over 65s, there are over 83,000 people with dementia, according to Alzheimer Scotland. It is estimated that currently 44 million people have dementia worldwide, with the total set to increase to 76 million by 2030, according to Alzheimer’s Society in the UK. They also estimate that dementia costs the UK economy £23bn a year which is more than cancer, stroke and heart disease combined.
One of the prominent problems for people with dementia is that they can get lost or become disorientated. For this reason, GPS trackers are used to locate patients. However, accurate location and indoor tracking is not feasible with GPS-based systems since satellite signals cannot work indoors nor do they offer sub-meter accuracy. The main challenge, therefore, is to achieve monitoring from a remote location of the status of people with dementia in terms of motion behavioural status in real-time. Knowing the accurate position of the person with dementia is therefore not only important for safety, but also for classification of ‘episodes’ of dementia for better management of the condition. Current techniques for indoor tracking use RF (radio frequency) sensors and these are not accurate enough, as they usually provide accuracy in the order of several meters. Hybrid systems which have additional sensors such as accelerometers and heading sensors offer somewhat improved accuracy at the expense of additional complexity since all these multiple sensors are then required to work in a coordinated fashion.
We propose a solution that is: lightweight, easy to install, relatively cheap to deploy and run. Our solution utilises LED (Light Emitting Diode) lighting, which is currently being installed in many homes due to its rapid drop in price, excellent colour rendering, LED longevity and LED energy efficiency. The solution works by utilising the fact that LEDs blink rapidly (this is imperceptible to a human eye) and that this blinking can be modulated to transmit data in the range of tens of megabits per second. We therefore propose using the LEDs to convey position information to a sensor located on the person. Due to the propagation of light properties, it is possible to achieve better precision in localization than with RF e.g. centimetres rather than metres.
Project Leader: Sinan Sinanovic
Project Researchers: Lynne Baillie, Wasiu Popoola, Roberto Ramirez Iniguez, Funmilayo Ogunkoya
Project Collaborators: North Glasgow Homes Housing Association
Project Funder: Digital Health Institute