Dr. ing. Maarten Sierhuis RIACS/USRA NASA Ames Research Center Moffett Field, CA msierhuis@mail.arc.nasa.gov http://www.agentisolutions.com Human activity behavior and gesture generation in high-fidelity VR training systems for Mars and Space Station operations. Abstract: In recent years, interest in collaborative virtual agents (bots) has increased due to the fact that most VR applications require close interaction with users and other systems. Although we have started from a different need, namely understanding the way people work, collaborate, and communicate in their environment (i.e. their work practice), we have developed an agent-based simulation environment (Brahms) that deals with a number of the same issues as the intelligent agent community is now addressing. Recently, we have started a research project (BrahmsVE) in which we are using Brahms as a multi-agent human-behavioral modeling language to develop bots in virtual worlds based on models of human behavior. The ultimate objective is to develop (uploadable) just-in-time VR training applications for human crews on Mars. Such applications need to include models of human behavior and work practice in order to accomplish a "life-like" interaction between astronauts (interacting as avatars) and bots in a virtual world. We claim that we need a closely coupled integration of an agent-based activity subsumption architecture for modeling human behavior with a low-level VR gesture generation capability, to develop a high-level behavioral VR bot-language. In this talk I will discuss our work over the past three years and discuss our Brahms/Adobe Atmosphere integration work. Keywords: bots, multi-agents, virtual worlds, simulation, work practice, human behavioral modeling. Speaker Bio: Dr. ing. Maarten Sierhuis is a senior researcher at the Research Institute for Advanced Computer Science (RIACS), an institute of the University Space Research Association (USRA), located at NASA Ames Research Center, Moffett Field, CA. His research focuses on work practice modeling and simulation and development of multi-agent systems in NASA domains, using the Brahms environment. Maarten’s research interests is in multi-agent languages for simulations and multi-agent systems, applied to the domains of work practices of the Apollo astronauts on the Moon, the astronauts onboard the International Space Station, mission operations for robotic missions, as well as human-robot interaction for human missions to Mars. He holds an engineering degree in Informatics, with a focus on knowledge modeling and expert systems, and a doctoral degree from the department of Social Science Informatics at the University of Amsterdam in The Netherlands, with a focus on modeling work practice.