Autonomous Agents: When the Mailbox Remains Empty
Résumé
An agent can send a message but never receive any response, which leads to handling problems. We name this issue the "Empty Mailbox Problem". The causes can lie in low level layers, for instance in communication links, but also in the higher layers, for instance in the behavior of autonomous agents, which can independently choose not to respond. Leaving the agent developer with the difficult task to determine what is to be done in such cases. The proposed solution presented in [8] consists in introducing a new performative and its associated protocol, producing a generic way to handle the empty mailbox problem. This paper refines the work made for this solution and focuses on the integration of this solution into interaction protocols.