The median age of European people is currently the highest in the world, and the proportion of people aged 65 or older is projected to increase by 45% in Europe in the next 20 years. As a consequence of ageing, a growing portion of people is at high risk of experiencing non communicable diseases, frailty, and social exclusion, and may need for long-term care, including nursing at home or frequent hospitalization. Of course, the inability of living independently may not only spoil the quality of life of elderly people and of those caring for them, but will also challenge the sustainability of the entire health system. Hence, it is urgent to devise innovative methods and tools for early detecting the onset of health issues and prolonging independent living in the senior population.
DomuSafe is an interdisciplinary project. The goal is to support active and healthy aging and early diagnosis of age-related conditions by continuously monitoring the activities carried out by the elderly at home, in a sustainable, unobtrusive and privacy-conscious manner. DomuSafe will rely on inexpensive wireless sensor technologies to acquire low-level information about the individual’s actions, and will devise innovative artificial intelligence methods to recognize activities and abnormal/unhealthy behaviors based on sensor data. Differently from most existing approaches, we will investigate unsupervised activity recognition method, in order to avoid the acquisition of activity training sets, which is a costly effort and violates individual’s privacy. Our method will rely on ontological and probabilistic reasoning to exploit the semantic correlations among activities and observed sensor events. The sensing infrastructure will be mainly based on cheap Bluetooth low-energy beacons with embedded accelerometer, which will be attached to furniture, devices and objects of interest to monitor their usage. The devised methods will be implemented in a prototype system, and evaluated both with existing real-world datasets, and with a set of patients recruited by the medical partner.
Funded by Region Sardinia under grant “Capitale Umano ad Alta Qualificazione“. Principal investigator: Daniele Riboni. April 2017 – April 2019.