No abstract

No abstract

This paper presents an argumentation based interaction within a multi-agent system designed for addressing the special requirements of accessibility content and services provisioning in an ambient intelligence context. We describe an architecture with multiple intelligent agents on the user's nomad device that address the ambient intelligence issue and cooperate with a family of dedicated personal assistance agents for each type of impairment, who are located on the server side. These agents provide the best consistent solution for people that have a combination of impairments through an argumentation based distributed decision making process. In this paper, we focus on presenting how we applied a particular argumentation framework for modelling communication and distributed decision making issues among a coalition of intelligent experts. The latter are looking for a compromise on the asked infomobility services, integrating different, very often conflicting, points of view due to the simultaneous existance of several impairments.

The HERA project, run by a consortium of researchers from academia and industry, applied the agent systems engineering methodology (ASEME) to develop a multi-agent system for a real-world application for ambient assisted living.services and applications to people: autonomy, distribution, adaptation, and responsiveness are key characteristics of AmI computing components. 2 Researchers working on the Home sERvices for specialised elderly Assisted living (HERA) project (w3.mi.parisdescartes. fr/hera), supported by the AAL Joint Programme (www.aal-europe.eu), built an AAL system to provide cost-effective specialized assisted living services for elderly people suffering from mild cognitive impairment (MCI), mild/moderate Alzheimer's disease (AD), and other diseases (diabetes, cardiovascular) with identified risk factors, aiming to significantly improve the quality of their home life and extend its duration. To address these needs, HERA provided the following main categories of services:• cognitive exercises, • passive communication,• pill and exercise reminders,• reality orientation (date and time), and E ngineering an ambient assisted living (AAL) information system using agent technology is a nontrivial task. 1 AAL systems combine the application domains of ambient intelligence (AmI) and home-based e-health. AmI refers to environments empowered with computing capability that deliver M u l t i -A g e n t S y S t e M S MAY/JuNE 2015 www.computer.org/intelligent Figure 4. The Personal Assistant (PA) role in the system roles model (SRM). The liveness property contains the formulas.