Abstract—— This discussion paper presents preliminary research in the field of Ambient Assisted Living systems. After a brief review of the existing literature, we argue that the research in this field is following a wrong paradigm: rather than creating technology to replace humans in assisting the elderly and disabled people, we should use technology to foster human interactions and enable more caregivers to get involved in this activity. We propose a prosumerist approach in designing AAL application, wherein producers and consumers of care may switch roles, and we outline an example of system designed according to this new paradigm. We also describe a method to solve the problem of activity recognition using low cost sensors and distributed neural networks. Index Terms: Ambient Assisted Living, Ambient Intelligence, distributed neural networks, social networks I. THE “MODERN TIMES” PARADIGM IN DESIGNING AMBIENT ASSISTED LIVING SYSTEMS S the population in developed countries continues to grow older, a great number of technical solutions attempt to respond to the rising demand for medical assistance and home care for the elderly or other persons with various cognitive, sensory, or motor disabilities These solutions received different names: Ambient Assisted Living (AAL), Ambient Intelligence (AmI), Home Care Systems (HCS), Smart Homes, Pervasive Healthcare, Telemedicine, e-Health, Smart Environments, etc., but all of them aim “to extend the time people can live in a decent way in their own home by increasing their autonomy and self- confidence, the discharge of monotonously everyday activities, to monitor and care for the elderly or ill person, to enhance the security and to save resources” ([1]). This general definition of AAL leads to the simple equation: AAL = Person in need + Assistive technology But, due to the extreme diversity of individuals in the target group and their specific needs, the proposed solutions are also diverse, hard to classify, and difficult to reproduce on a large scale. Manuscript received June 8, 2012. I. Susnea is with the Department of Computers and Information Technology, University “Dunarea de Jos”, of Galati, Romania (e-mail: ioan.susnea@ugal.ro) G. Vasiliu, is with the Department of Automation and Applied Informatics, University “Dunarea de Jos”, of Galati, Romania (e-mail: grigore.vasiliu@ugal.ro). S. Spiridon is with the Regional Center for Resources and Educational Assistance (CJRAE) Galati, Romania (e-mail:simonaspiridon@yahoo.com) How efficient are these solutions? Is a multi-million dollar smart home more efficient than a two dollar cane or a pair of crutches in terms of increasing the autonomy and self confidence of the elderly? Do they really reduce the burden of care for the increasing number of patients and the overall economic and social costs of assistance? In an attempt to answer these questions, we reviewed in part, the vast literature on these topic, including some recently funded AAL research projects listed in [2] and [3]. We noticed that many of the recent solutions in AAL follow the design paradigm defined by the feeding machine presented in the famous Chaplin’s movie “Modern Times”. Despite all major advances in technology, the assisted person is still an “object”, and – in a pure industrial revolution approach - the design focus is on increasing the work productivity of the caregivers. We also noticed that there is an important neglected resource in AAL: the other people. Therefore, we propose an amendment to the equation that defines AAL systems: AAL = person in need + other humans + technology In our vision, the AAL technology should not attempt to replace humans, but to enable and foster complex interactions between them, in order to attract more caregivers in human centered AAL systems. On the other hand, designing machines capable to replace humans is not an easy task. Therefore, some of the resulting solutions, though technically feasible, tend to be extremely complex, expensive, and totally impractical. In this context, the purpose of the research described in this paper is thereefold: (i) emphasize the need for a paradigm shift in the design of AAL systems, (ii) outline a solution to reduce costs of some AAL applications, and (iii) propose the concept of prosumerism ([4]) in caregiving and outline a possible implementation of such system. Beyond this introduction, this paper has the following structure: - Section II is a brief review of the state of the art in AAL and related fields, aimed to identify the current major design approaches in AAL - Section III proposes a simple solution for the problem of activity recognition in Ambient Intelligence systems, based on distributed neural networks, with the aim to reduce costs and improve reproducibility of such systems. - Section IV proposes a new paradigm in designing AAL systems, and outlines the structure of a system built according to this new paradigm. - Section V is reserved for conclusions. A Discussion Paper on Designing Ambient Assisted Living Systems Ioan Susnea, Grigore Vasiliu, and Simona Spiridon A