Influencing Actions-Related Decisions Using Soft Computing Approaches

Robotic Elder Care

Robotic aids for the disabled and elderly is a growing area of research (Goth, 2011; Jacobs & Graf, 2012). This is strongly motivated by the rapidly increasing population in this age group. Japan has several active research programs (Kanda, Nabe, Hiraki, Ishiguro & Hagita, 2008; Severinson-Eklundh, Green & Huutenrauch, 2003) as 22% of their population is 65 or older and quickly expanding in the coming decades. In the United States the National Science Foundation has established the Quality of Life Technology Center at Carnegie Mellon and University of Pittsburgh.

For this paper we will illustrate our approach using a scenario in which a robotic assistant (RA) in the future can provide various forms of assistance in rehabilitation and even daily living requirements of the elderly. Often elderly residents of a nursing home may be somewhat addled or recalcitrant. So it is very important that a robotic assistant be able to communicate to persons with whom they are trying to persuade about issues such as medications or rehab activities. In particular social aspects of the assistants' interaction have to be considered (Bemelmans, Gelderblom, Jonker & de Witte, 2012; Klein, Gaedt & Cook, 2013). We will describe how this communication can be modeled with soft computing techniques and recognize the social aspects of such interactions.