Enhanced Deliberation in BDI-Modelled Agents

Applications that operate in highly dynamic environments must deal with real-time changes of circumstances in order to be consistent and coherent. In this work, we propose an extended deliberation cycle for BDI-modelled agents that takes advantage of environmental events to coordinate the deliberation process. Our argument is that is it possible to optimise the deliberation performance by exploiting features of windows of opportunity. The proposed model let us define precisely the expected application behaviour that promotes the balance between reactive and proactive applications, leading to solutions with improved computational performance.