A framework of awareness for artificial subjects

We review the concepts of environment-and self-models, semantic interpretation, semantic attribution, history, goals and expectations, prediction, and self-inspection, how they contribute to awareness and self-awareness, and how they contribute to improved robustness and sensibility of behavior. Researchers have for some time realized that a sense of “awareness” of many embedded systems' own situation is a facilitator for robust and dependable behaviour even under radical environmental changes and drastically diminished capabilities. This insight has recently led to a proliferation of work on self-awareness and other system properties such as self-organization, self-configuration, self-optimization, self-protection, self-healing, etc., which are sometimes subsumed under the term “self-*”.