Modelling and verifying BDI agents with bigraphs

• We show BDI agent programming languages can be encoded in Milner's Bigraphs. prove the encoding of a superset AgentSpeak, e.g. CAN language, is faithful. provide verified executable semantics for language programs. verify program properties through model checking. This foundation to enable advanced reasoning, probabilistic verification. The Belief-Desire-Intention (BDI) architecture popular framework rational agents; existing verification approaches either directly encode simplified (e.g. lacking features like failure recovery) into frameworks Promela), or reason about specific implementations . take an alternative approach and employ bigraphs as modelling fully featured Conceptual Agent Notation (CAN)—a AgentSpeak featuring declarative goals, concurrency, recovery. syntax Can agents, give rigorous proof that Verification based on use mainstream software tools including BigraphER, small case study verifying several Unmanned Aerial Vehicles (UAVs) illustrates action. foundational step will more reasoning such plan preference, intention priorities trade-offs, interactions with environment under uncertainty.