Proceedings, Komputer dan Sistem Intelijen (KOMMIT2004) Auditorium Universitas Gunadarma, Jakarta, 24 – 25 Agustus 2004 ISSN : 1411-6286 Understanding User Activity in Distributed Intelligent Environment 9 UNDERSTANDING USER ACTIVITY IN DISTRIBUTED INTELLIGENT ENVIRONMENT Teddy Mantoro Department of Computer Science, Australian National University, ACT 0200, Australia teddy.mantoro@anu.edu.au Abstract. User activity is the most complex part of the context awareness mechanism in an Intelligent Environment. Unfortunately, the computer environment changes overtime. The problem confronting the user is how to enable transparent distribution computing to continue operation across a changing circumstance in a seamless manner and how the environment recognises user activity. This paper describes an approach to determine and recognise user activity based on user location. A sa part of user activity, a mobile access point for a ‘guest’ user to access resources in an unfamiliar domain is also introduced. This work is based on DiCPA architecture, a distributed context processing architecture for smart environments. We show how the scalable distribution of context and context information is managed to characterise user activities. Keywords: User Activity, User Location, Context Awareness, Pervasive Computing, Ubiquitous Computing, Active Office, Intelligent Environment 1. Introduction As we aware in the last decade computer environment changes over time and at the same time user needs also become more complicated. The changing includes user profile - from a single profile for a single device to many profiles and many devices for a user -, user location, situation and resources as well as the network configuration. This is one of the impacts that enable technology from wired network to wireless network, from Local Area Network to Personal Area Network or to Wide Area network. The problem is how to enable transparent distributed computing to continue operation across a changing circumstance in a seamless manner. How to exploit the changing environment that is aware of the context of the location use, the collection of nearby people and objects, the accessible devices and the change to those objects over time. Moreover, how smart environments have the capability to assist and help people with a variety of activities by detecting a users’ current state/context to determine what actions to take based on the context. Recently, scientists in the Intelligent Environment (IE) area are researching ways to make embedded computing and ubiquitous computing work better for people by creating and equipping a smart environment, such as an active home or an active office, with technologies that can identify user needs and meet them speedily, efficiently and unobtrusively. The goal of context aware computing in general is to make user interacting with the computer easier in the smart environment where technology is spread throughout (pervasive), computers are everywhere at the same time (ubiquitous) and technology is embedded (ambient) in the environment. It need not difficult, tedious or need hard learning to the user. It should potentially be safe, easy, simple and enable new functionality without need to learn a new technology. It provides relevant information and a simple way for a user to deal with it. The end uses of such technology could simply be convenience in an active home, knowing when the occupant wakes up and what radio station he likes to listen to without waking up the rest of the house. On the other hand it could be life saving by detecting when he had fall and needs medical help. Our approach is to locate a person within an environment using the wireless connections in devices that they normally carry for other purposes, for example, a mobile phone, PDA or a laptop