The human visual system consists of a high-resolution fovea and a low resolution periphery. This arrangement requires us to use eye and head movements to bring objects of interest, detected with low-resolution peripheral vision, to the fovea for more detailed analysis. We do this successfully because humans have evolved effective eye movement strategies for performance of everyday tasks. In addition, individuals develop special strategies as they become experienced, and eventually expert, in the performance of specific tasks, for example driving, painting or working on an assembly line. Robotic visual systems must often choose between speed and depth of processing. Such systems can clearly benefit from human-like strategies for allocating their limited resources. The purpose of this research is to study eye movements of human subjects as they perform tasks for which robotic systems are being developed. The goal is to identify human scanning strategies that may improve the performance of robots in similar tasks and then verify their effectiveness by adapting these strategies to robots being developed.