Design and evaluation of a device worn for fall detection and localization: Application for the continuous monitoring of risks incurred by dependents in an Alzheimer’s care unit Yoann Charlon a,b,⇑ , Nicolas Fourty c , Walid Bourennane a,b , Eric Campo a,b a CNRS, LAAS, 7 avenue du Colonel Roche, F-31400 Toulouse, France b Univ de Toulouse, UPS, UTM, LAAS, F-31400 Toulouse, France c LCIS, Grenoble-INP, Université Pierre Mendès, 50, rue Barthélémy de Laffemas, BP54, F-26902 Valence, France article info Keywords: Elderly Care unit Smart monitoring system Wireless sensor network Fall detector Indoor localization abstract The Homecare project, which is part of a research project funded by the French National Research Agency (ANR), aims to define a new multi-sensor monitoring system for the elderly with cognitive disabilities in a care unit. Two subjects were recruited to participate to experimental trials. The main objective of this project is to design and test a complete monitoring system at a real site. It is a new clinical and technical approach which is complex to implement: Homecare is intended to propose a possible technical solution, demonstrate its feasibility and illustrate its use working at a protected site. The system consists of a motion sensor network deployed on the ceiling to monitor motion and an electronic patch worn by the subjects to identify them and detect falls. In order to locate tagged subjects inside the care unit, a net- work of anchor points is used. From these positions and movement data, an analysis algorithm detects an abnormal situation and alerts the nursing staff in real time. A Web application allows the medical staff to access movements and alarms. The complete monitoring system has been functioning for several months and continuously monitors two patients around the clock. In this paper, we present the implementation of the system, the method of localization inside the care unit, and the characterization of the fall detector, and we show certain results relating to activity data. Ó 2013 Elsevier Ltd. All rights reserved. 1. Introduction The aging population in developed countries has highlighted several economic and societal problems over the last decades. Indeed, the care of dependent people has emerged as a major economic and ethical issue. The situation in Europe is critical. For instance, France has more than 700,000 dependent people and this figure is predicted to increase by 2% a year until 2040. By 2050 (Duée & Rebillard, 2004), approximately 1.5 million dependent people will require the assistance of another person. Similar demo- graphic changes are taking place in most European countries, the USA and Japan. Overall, approximately 20% of the world population will be at least 60 years old by 2050 (Chan, Estève, Escriba, & Campo, 2008). Health technologies represent an important market and develop products with high added value. The overall market is estimated at €185 billion and its annual growth is approximately 6–7% (Pammolli et al., 2005). The French market was estimated at €6.7 billion in 2005 (Pammolli et al., 2005). It includes medical devices, technical assistance, medical benefits and social/nursing facilities. Market studies suggest that autonomy and homecare has a potential value of €4 billion annually in France and will pro- vide 50,000 new jobs, a figure which could be multiplied by three to four by 2050 (ALCIMED, 2007). In a context of economic crisis, sensor networks and miniatur- ized devices are offering new technological solutions in the care of dependent persons. Research and development in smart moni- toring systems, the purpose of which is to look after the elderly liv- ing at home or in institutions, has gathered considerable interest in academia and certain industries (Chan et al., 2008; Gentry, 2009). Such systems permit real time monitoring of movements and gen- erate emergency alerts (abnormally rapid heartbeat, falls, etc.). They generally use sensor networks disseminated in the environ- ment or worn by the person (e.g. on clothes or the skin). Some of these systems enable the monitoring of activities of daily living (ADLs) Fleury, Vacher, & Noury, 2010; Cerny, 2010; Helal et al., 2005; Kientz et al., 2008: life habits, movements, immobility, and restlessness. In order to monitor ADLs, some systems use sound or video (Kidd et al., 1999; Krumm et al., 2000; Riedel, Venkatesh, & Liu, 2012; Yamazaki, 2006). They have often been criticized and described as an intrusion into personal privacy. Some research 0957-4174/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.eswa.2013.07.031 ⇑ Corresponding author. Tel.: +33 661451045. E-mail addresses: ycharlon@laas.fr (Y. Charlon), nicolas.fourty@iut-valence.fr (N. Fourty), wbourenn@laas.fr (W. Bourennane), campo@laas.fr (E. Campo). Expert Systems with Applications 40 (2013) 7316–7330 Contents lists available at SciVerse ScienceDirect Expert Systems with Applications journal homepage: www.elsevier.com/locate/eswa