Bees in clutter: observing flight strategies to inspire visually guided autonomous navigation

Abstract Insects have long served as inspiration for robotic systems for various applications. Among them, bees have emerged as promising models for the discovery and implementation of strategies for aerial navigation in autonomous robots. This is motivated by bees’ ability to perform robust visually guided behaviors to navigate complex spatial environments safely. The specific flight behaviors that these agile organisms perform when negotiating visual clutter are particularly relevant, as modern autonomous vehicles operate in similar cluttered environments, where parsimonious sensing and robust control are necessary. This review focuses on synthesizing discoveries from behavioral studies of bee flight in clutter, alongside complementary advances in implementing bio-inspired navigation strategies for robotic platforms. The sensory mechanisms and salient properties of crucial visual inputs, such as optical flow, which aid bees in navigating clutter, are discussed from a functional standpoint. Prevailing hypotheses on the strategies to accomplish key behaviors, including speed regulation, collision avoidance, and gap traversal, are described. To facilitate the translation to robotics, an outline of control laws for producing bee-like navigation behavior in representative cluttered environments is presented. Finally, we collate work from the past decade on bee-inspired robotic control and highlight avenues for future research to realize embodied intelligence in robotics. Through our analysis and survey, we identify open questions and research directions to deepen our understanding of the neuroethology of flight behavior in bees, as well as recent trends in the development of bio-inspired robotics research that operates at the interfaces of biology, engineering, and robotics.