A Geometric Control Framework for Enabling Behavior-Based Planning and Locomotion of Undulatory Robots

点击翻译按钮获取中文摘要

Equipped with a slender and over-actuated body, snake-like robots have the ability to maneuver through complex environments. These platforms have been heralded for their locomotive advantages when navigating through rubble, tight spaces, and other areas hard to reach with standard robots. Yet, without a reliably systematic feedback control framework, snake-like robots cannot make the transition from teleoperation to full autonomy. This award will support advances in the dynamic modeling of snake locomotion from fundamental mechanical principles. These models are important for controlling robotic snakes, and other undulatory systems, and will remove a key obstacle to autonomy. The award will also support fundamental research into the role that scales play in snake locomotion, including the creation of robotic snake skin that will enable snake-like robots to realize the locomotive advantages of actual snake scales. The end goal of the project is to demonstrate autonomous snake operation through an unknown environment. Doing so will bring robotic snakes significantly closer to being deployable for search and rescue, inspection, and operation in hazardous environments. The research findings will also apply to similar biologically-inspired robots, such as swimming and flapping-wing flying robots. The popularity of robotics will be leveraged to create engaging educational and outreach activities in promotion of engineering mathematics.Snake-like robots are high-dimensional, articulated limbless robots that exploit control over a continuous morphological feature to achieve locomotion in a variety of ways. Science has exposed the physical principles underlying locomotion, while engineers have reproduced the fundamental mechanisms and can teleoperate these robots. However the creation of actionable feedback control policies is not fully resolved. What is missing is the physical and mathematical formulation that bridges the gap from achievement of undulatory gaits to the controlled execution of motion along a planned path. To address the gap, this project will design, and demonstrate the performance benefits of, biologically inspired scale fabrication for robotic snakes. The designed scales will reproduce the structural and anistropic friction properties of snake scales. Further, the research will use spatio-temporal averaging to derive reduced control equations for undulatory robotic systems where the body's internal degrees of freedom are modeled in the continuum. Achieving these two objectives will resolve a long-standing achievement gap between biologically inspired, undulatory robots and traditionally engineered robots.