This paper introduces a notion of forward thinking agents. Recent research in situation awareness and anticipatory behavior provides useful insights and the relevant cognitive underpinnings that can be used to design adaptive software agents that capitalize on the environmental cues and their repercussions in the context of motivational autonomy. In particular, we are interested in building software agents that exhibit non-trivial anticipatory behaviors - exploiting knowledge about the current situation and the future to improve their current behavior. Some of the key issues pertaining to the realization of forward thinking, anticipatory behavior and situation awareness in agency will be briefly discussed. A conceptual extension to the generic BDI framework is also presented as part of our initial efforts in tackling some of these issues.
|Title of host publication||Agents and Computational Autonomy|
|Subtitle of host publication||Potential, Risks, and Solutions|
|Editors||M Nickles, M Rovatsos, G Weiss|
|Number of pages||12|
|Publication status||Published - 2004|
|Name||Lecture Notes in Computer Science|