Program

09:00-10:45 An Introduction to agent technology - Frances Brazier and Jeff Kephart (slides)

11:15-12:00 Invited talk - Michael Huhns (slides)

13:30-14:00 Contributed paper

Evangelos Pournaras, Martijn Warnier and Frances Brazier -
Using intelligent agents for self-adaptation and self-optimization of
energy consumption in power networks (paper), (slides)

14:00-14:45 Invited talk - Van Parunak (slides)

15:00-16:00 Panel and discussion (slides)

Self-management of complex systems is core to both the Autonomic Computing and the Software Agent communities. In both paradigms, individual autonomous entities manage their own behaviour and their interactions with the environment and other autonomous entities in accordance with their individual goals based on their local perception of state. These entities may negotiate with one another, and monitor and manage the resulting agreements. They may form dynamic virtual organizations that manage their collective behaviour in interaction with other such organizations. They may avail themselves of integration, repair and other services provided by directories, brokers and sentries, which themselves may be autonomous.

Over the course of many years, the software agents community has developed and explored architectures, technologies and standards that support these aspects of agent behaviour, and have demonstrated in multiple contexts agents and multi-agent systems that exhibit autonomy, goal-directed adaptive behavior, proactivity, reactivity, situated-ness, and an ability to learn. The relatively younger field of autonomic computing seeks to build computing systems that exhibit these same properties and capabilities, but with few exceptions has failed to tap into the rich body of knowledge developed by the agents community. Some authors have suggested that autonomic computing may be the long-sought "killer app" for agents.

It seems clear that the Agents and Autonomic Computing communities have much to gain from a closer association with one another.

The aim of this workshop is to:

1. explore the potential of the agent paradigm, architectures, models and technology for autonomic computing;
2. identify the specific challenges of autonomic computing that would require extensions to the agent paradigm and current agent technologies;
3. discuss the feasibility of a joint research/development agenda.

Topics of interest include, but are not limited to:

• (Meta-)Architectures for agents and multi-agent systems
• Coordination, scheduling, planning, clustering
• Learning algorithms
• Adaptivity, situatedness
• Emergent behaviour, emergent configurations
• Service agreements
• Negotiation
• Large scale simulations/emulations
• Mobility
• Legal implications of self-management/autonomy in networked systems
• Accountability, verification and validation
• Reliability, Integrity and Security

ORGANISING COMMITTEE

PROGRAMME COMMITTEE

IMPORTANT DATES

WORKSHOP FORMAT

This one-day workshop will include 4-6 invited talks, paper presentations, a forum/panel discussion and time for discussion. The workshop is open to all ICAC participants for no additional fee.

PAPERS

Papers 4-6 pages in length in the standard IEEE two-column conference proceedings format (style files)
Papers submission instructions.

Workshop proceedings will be distributed during the ICAC conference.