SB-MASE is aimed at students and researchers in AI, for simulation and control of agents and robots. It uses task based decomposition and the subsumption architecture as its core control mechanisms and includes numerous possibilities for adding plug-ins.