Abstract
Purpose: This article aims to the evaluation of a prototypal assistive technology for Alzheimer's disease (AD) patients that helps them to remember personal details of familiar people they meet in their daily lives. Method: An architecture is proposed for a personal information system powered by face recognition, where the main AD patient's interaction is performed in a smart watch device and the face recognition is carried out on the Cloud. A prototype was developed to perform some tests in a real-life scenario. Results: The prototype showed correct results as a personal information system based on face recognition. However, usability flaws were identified in the interaction with the smart watch. Conclusions: Our architecture showed correct performance and we realized that it could be introduced in other fields, apart from assistive technology. However, when being targeted to patients with dementia some usability problems appeared, such as difficulties to read information in a small screen or take a proper photo. These problems should be addressed in further research.