Ambient Intelligence Through Agile Agents - Ambient Intelligence for Scientific Discovery - page 285

Ambient Intelligence Through Agile Agents

Gregory M.P. O'Hare, M.J. O'Grady, R. Collier,
S. Keegan, D. O'Kane, R. Tynan, and D. Marsh

University College Dublin (UCD), Ireland
[e-mail]

1 Introduction

The vision of ambient intelligence is one where the populas is supported in the
conductance of their everyday lives through the pro-active, opportunistic support
of non-intrusive computing devices offering intuitive interaction modalities.
This chapter advocates the adoption of mobile intentional agents as a key
enabler in the delivery of ambient intelligence. Ambient computing, as an ideal,
demands levels of functional attainment that have hithertofar not been realized.
Ambient applications demand that the computing application be subsumed into
the everyday context in an unobtrusive manner with interaction modalities which
are natural, simple and appropriate to both the individual user and their asso-
ciated context.
Ambient systems need to address some key issues:
- Recognition and accommodation of the diversity of devices that contribute
to the organic nature of the ambient and ubiquitous computing nervous
system;
- The need for personalization and system adaptivity;
- An understanding of the dynamics of context;
- Provision of support for collaboration and cooperation between distributed
ambient system components;
- Delivery of systems that exhibit autonomic characteristics, yielding self man-
agement and self healing capabilities;
In addressing these core issues we commission an intentional agent based
approach. Specifically we adopt a Belief-Desire-Intention (BDI) agent model. In
the delivery of such agents we utilize the Agent Factory system. The Agent Fac-
tory (AF) System is an environment developed in part by two of the authors,
which supports the rapid fabrication of agent-based applications. Agent Factory
provides an integrated environment for the development of agent based systems
providing a methodological framework together with an accompanying software
membrane which provides a cohesive and integrated tool set which supports the
various stages in the design, specification, implementation, debugging and visu-
alization of agent behavior. Detailed descriptions of Agent Factory are provided
elsewhere in the literature [3, 4, 27, 28]. Agent Factory supports Weak Migration
where only the agent's object state and code is captured. Upon migration the
system calls a known entry-point in the code to restart the agent on the new
machine.