Metaheuristics and Cognitive Models for Autonomous Robot Navigation

Although autonomous mobile robots have become more commonplace, methods for them to navigate in the real-world are not yet perfected. This survey describes two heuristic-based approaches, metaheuristics and cognitive models, that have been used in this domain under separate but concurrent development. Metaheuristics are broadly-applicable, heuristic-based algorithms inspired by natural processes. They use one or more heuristic strategies to search for solutions. Alternatively, studies have shown that people adaptively use one or more strategies during navigation. Cognitive models are used to computationally encode human behaviors. This paper details both approaches, reviews the literature on their application to autonomous robot navigation, and discusses their advantages and disadvantages. Finally, to take advantage of algorithmic efficiency and human behaviors, the conclusion proposes the integration of these two approaches.