NRI/Collaborative Research: Improving the Safety and Agility of Robotic Flight with Bat-Inspired Flexible-Winged Robots

点击翻译按钮获取中文摘要

Bat flight, perhaps the most advanced and efficient form of animal flight, has long been a source of inspiration for roboticists and biologists alike. This National Robotics Initiative (NRI) collaborative research award supports research aimed at understanding and reproducing the unparalleled agility and resilience of bat flight. Biological studies of bats (their structure, muscle movement, and flight dynamics) will drive the engineering development of mathematical models of robotic flight and the eventual design and implementation of a prototype 30-80cm bat-like robot. The physical flight capabilities of the robot will be augmented with perception and reasoning abilities, with the aim of providing support for construction site activities such as site monitoring, inspection, and general surveillance of the work site to provide image data to enhance situational awareness of human workers. The research involves several disciplines, including biology, aerodynamics, robotics, control systems engineering, and construction engineering.Aerial robots have nowhere near the agility and efficiency of animal flight, especially in complex, constrained environments. This is not surprising since even the simplest winged robots have complex flight dynamics that pose significant challenges for modeling, design, and control. In the case of bat-inspired robots, these difficulties are exacerbated by the use of under-actuated mechanisms driving wings constructed from flexible membranes. This project will combine biological and engineering research to address these problems. Biological research on the kinematics of bats and their flight will provide a basis for mechanical designs. To control the robot, agile motion planning and flight control algorithms will employ motion primitives that are derived from biological investigation of the dynamics of bat flight. Conversely, models obtained from biological studies will be validated by experimental investigations using the prototype robot, enabling iterative refinement of reduced-order models and control algorithms. Ultimately, the robots will be equipped with sensing systems and planning algorithms, to facilitate localization, mapping, inspection and surveillance at construction sites.