We review recent research in robotics, neuroscience, evolutionary neurobiol
ogy, and ethology with the aim of highlighting some points of agreement and
convergence. Specifically, we compare Brooks' (1986) subsumption architect
ure for robot control with research in neuroscience demonstrating layered c
ontrol systems in vertebrate brains, and with research in ethology that emp
hasizes the decomposition of control into multiple, intertwined behavior sy
stems. From this perspective we then describe interesting parallels between
the subsumption architecture and the natural layered behavior system that
determines defense reactions in the rat. We then consider the action select
ion problem for robots and vertebrates and argue that, in addition to subsu
mption-like conflict resolution mechanisms, the vertebrate nervous system e
mploys specialized selection mechanisms located in a group of central brain
structures termed the basal ganglia. We suggest that similar specialized s
witching mechanisms might be employed in layered robot control architecture
s to provide effective and flexible action selection.