Action and Perception in Agent Programming Languages: From Exogenous to Endogenous Environments

The action and perception models adopted by state-of-the- art agent programming languages – in the context of Multi-Agent Sys- tem (MAS) programming – have been conceived mainly to work with exogenous environments, i.e. physical or computational environments completely external to the MAS and then out of MAS design and pro- gramming. In this paper we discuss the limits of adopting such models when endogenous environments are considered, i.e. computational en- vironments – often referred also as application environments – that are designed and programmed by MAS developers as a first-class abstraction to encapsulate functionalities useful for agent individual and cooperative activities. In the paper we describe an action and perception models for agent programming languages specifically conceived to be effective for en- dogenous environments and we discuss its evaluation using CArtAgO en- vironment technology. On the agent side, we focus our attention on pro- gramming languages based on the BDI (Belief-Desire-Intention) model, taking Jason, 2APL and GOAL as reference case studies.