Agents, games and HLA

Over the past decade, there has been a growing interest in utilising intelligent agents in computer games and virtual environments. At the same time, computer game research and development has increasingly drawn on technologies and techniques originally developed in the large scale distributed simulation community, such as the HLA IEEE standard for simulator interoperability. In this paper, we address a central issue for HLA-based games, namely the development of HLA-compliant game agents. We present hla_agent, an HLA-compliant version of the sim_agent toolkit for building cognitively rich agents. We outline the changes necessary to the sim_agent toolkit to allow integration with the HLA, and show that, given certain reasonable assumptions, all necessary code can be generated automatically from the FOM and the object class publications and subscriptions. The integration is transparent in the sense that the existing sim_agent code runs unmodified and the agents are unaware that other parts of the system are running remotely. We present some preliminary performance results, which indicate that the overhead introduced by the HLA extension is modest even for lightweight agents with limited computational requirements.