A survey of animal foraging for directed, persistent search by rescue robotics

Abstract

This paper reviews the animal foraging literature for insights into the directed, persistent search problem in rescue robotics, where a robot may have to move away from a target or out of sensing range in order to find a path to it. The conclusions from 32 papers on empirical studies of insects, fish and birds, mammals, and 20 studies exploring theoretical models are that: robots should divide the environment into patches, they should follow a win-stay search strategy within a patch, use either a sensory- or pattern-based search algorithm, should adapt the algorithm to travel distance, total search time, physical obstacles, and availability of sensory information, and persistence should have a fixed time limit unless sensory information is available. The survey is expected to enable rescue roboticists to create better, or improve existing, search strategies and algorithms.