Home // SMART 2012, The First International Conference on Smart Systems, Devices and Technologies // View article

Ambient Assistance Using Mobile Agents

Authors:
Amine Arezki
Eric Monacelli
Yasser Alayli

Keywords: Human Activity Recognition; Ambient Assistance; Human Tracking; Robot Agent.

Abstract:
In this paper, a method is presented to assure ambient assistance in an urban environment, using a mobile agent. The goal is to anticipate the assistance if needed. Therefore, the robot has to understand the human behaviour and a person’s needs. We will determine how to focus on a moving subject and using interaction for confirmation, in order to provide an assistance if needed. Therefore, the trajectory type concept is used to define the first step of analyses, which is called the approach step. Combining this step with the field information provided by the mobile agent will insure a certain type of assistance. In terms of observation, two different views are employed to detect assistance requirements, i.e., the Fix Intelligent Device and the Ambient Intelligent Devices, both communicating by Wi-Fi. The Fix Intelligent Device is a fix camera standing on a very top view allowing the detection of possible motions and classification of trajectories, using neural network. In our research, a Touch Ambient Intelligent Device is utilized, a mobile robot with three degrees of freedom, including a 3D camera (KinectTM) and a touch screen to interact with the subject. In this paper, the behavior of the mobile agent, receiving detected trajectories emitted by the Fix Intelligent Device, is presented and it is shown that the human intervention is only needed in critical cases.

Pages: 89 to 95

Copyright: Copyright (c) IARIA, 2012

Publication date: May 27, 2012

Published in: conference

ISSN: 2308-3727

ISBN: 978-1-61208-225-7

Location: Stuttgart, Germany

Dates: from May 27, 2012 to June 1, 2012