Detection of Behavioral Changes on Activities Daily Routine in a Smart Home using Heatmaps

With the rapid growth of the elderly population, there is a need to propose a solution to help elderly people remain at home and maintain their autonomy and healthy lifestyle while reassuring the relatives, caregivers, and healthcare professionals. The advances in pervasive computing technologies have opened up new opportunities such as smart homes to remotely monitors the elderly activities at home and detects any changes in their behavior. In this paper, we propose an unsupervised approach to detect any changes in the elderly daily behavioral routine. We use heatmaps as a way of representing the elderly daily behavioral routine, autoencoder to reduce the dimensional space of the features and a clustering algorithm to detects the different behaviors of the smart home resident. We evaluate our approach on a real-life dataset and a compare it to an existing approach. The results are promising and they suggest that changes in the elderly daily routine can be identified and presented graphically.