P. Kostkova (Ed.): eHealth 2009, LNICST 27, pp. 62–69, 2010. © Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering 2010 Detecting Human Motion: Introducing Step, Fall and ADL Algorithms Dries Vermeiren, Maarten Weyn, and Geert De Ron Artesis University College of Antwerp, Antwerp, Belgium dries.vermeiren@artesis.be, maarten.weyn@artesis.be, geert.deron@ieee.org http://www.e-lab.be/ Abstract. Telecare is the term given to offering remote care to elderly and vul- nerable people, providing them with the care and reassurance needed to allow them to keep living at home. As telecare is gaining research interests, we'll in- troduce a system which can be used to monitor the steps, falls and daily activi- ties of high risk populations in this paper. Using this system it is possible for a patient to rehabilitate at home or for elderly to keep living independently in their own house while they are still monitored. This leads to a huge cost reduc- tion in health services and moreover it will make patients satisfied for being able to live at home as long as possible and in all comfort. Keywords: MEMS, Step Detection, Fall Detection, ADL, Freescale. 1 Introduction Elderly people are the fastest growing segment of the population. Due to Europe's population pyramid, aging people are becoming a point of interest even faster than in the rest of the world. In 2035, one third of the Europeans will be more than over 65 years old, which will result in huge strains on health care services [5]. This will also cause a serious social and financial problem. Care centers will have to deal with a lack of rooms and cost reduction will become one of the most important objectives in public health services [5]. One of the possible solutions is to comply with the wish of elderly to keep living at home independently as long as possible. As we do this, an increasing number of high risk populations will be living alone at home. Therefore new advanced monitoring systems are gathering more research popularity. Not only for the elderly, but also for other high risk populations (people su_ering from illnesses such as epilepsy or Alzheimer or in the case of recent surgical intervention) will long- term monitoring become an issue. In this paper we will describe a system based on 2 tri-axial accelerometers to detect the Activities of Daily Living (ADL) of a patient and to detect its steps and falls. In the first part we will focus on the most appropriate position of the sensor on the pa- tient's body in order to receive clear signals of their movements. Furthermore we will describe different methods for detecting steps with techniques based on simple filters and thresholds or templates. Next to this we will also study the different methods to detect falls and a very basic ADL detection method is proposed. Subsequently we will