A self-aware system has the ability to learn and maintain knowledge of itself, its experiences, and interactions with its environment, on an ongoing basis. This permits reasoning and decision making to support effective, autonomous adaptive behaviour, in an uncertain and changing environment. This extended abstract describes recent work to develop the concept of computational self-awareness, a notion of machine self-awareness inspired by theories of human self-awareness, and that arises through computational processes. Recent work to develop conceptual reference architecture is high-lighted, and a key role for online learning is identified.
CODES/ISSS ’16, October 01-07 2016, Pittsburgh, PA, USA