Lévy Flight Foraging Hypothesis-based Autonomous Large-scale Memoryless Search under Sparse Rewards

Autonomous aerial robots are commonly tasked with the problem of area exploration, surveillance and search for certain targets or objects of interest to be detected and tracked. Traditionally, the problem formulation considered is that of complete coverage and thus - ideally - identification of all targets of interest. However, an important problem which is not often addressed is that of memoryless time-efficient search under sparse rewards that may be worth visited any number of times. An indicative application scenario relates to surveillance for moving and possibly camouflaged targets thus making map coverage an inherently memoryless process. In this paper we specifically address the largely understudied problem of optimizing the “time-of-arrival” or “time-of-detection” to robotically search for sparsely distributed rewards (detect targets of interest) within large-scale environments and subject to memoryless exploration. At the core of the proposed solution is the fact that a search-based Lévy walk consisting of a constant velocity search following a Lévy flight path is optimal for searching and identifying distributed target regions in the lack of map memory. A set of results accompany the presentation of the method, demonstrate its properties and justify the purpose of its use towards large-scale area exploration autonomy using aerial robots.