Self-modeling and self-awareness

Abstract

The purpose of this chapter is to discuss why self-aware systemsmust pay special attention to self-modeling capabilities, clarify what is meant by both strong and weak self-modeling, and describe some of the defining characteristics of selfmodeling. This chapter is also about self-management via run-time model creation by the operational system, explaining why systems need to build models at run time, what phenomena they need to model, and how they can build models effectively. A system that is expected to operate in a dynamic environment needs to be able to update and occasionally dramatically change its models to maintain synchrony with that environment. We describe several example systems, one rather extensively, to show how the notions apply in practice.