Using reconfigurable modular robots for rapid development of dynamic locomotion experiments.

摘要

In locomotion research, prototypes ranging from purely passive mechanical linkages to full-fledged autonomous mechatronic machines are built to validate locomotion principles and explore different morphologies. Being able to quickly build robotic prototypes has the capability to improve workflow, productivity, and innovation. Modular Robots, for instance, allow us to build robots quickly, and rapidly explore different morphologies.;We present the design and development of a Modular Robot system called CK-Bot. One of the major innovations of this system is a connection mechanism that allows the robot to be instantaneously reconfigured manually, while still maintaining a robust connection. We show the practical utility of rapidly building machines with modules in product design, and emergency response, but choose to focus on dynamic locomotion research. To show that this system can indeed be a useful tool for dynamic locomotion research, we use two of the prototypes and analyze their dynamic locomotion principles.;The first locomotion principle is a loop configuration that uses a sensor-based feedback controller to achieve dynamic rolling. The robot senses its position relative to the ground and changes its shape as it rolls. Using simulation and experimental results, we show ways in which the desired shape can be varied to achieve higher terminal velocities. One of our major findings is that more elongated shapes achieve higher terminal velocities than rounder shapes. We also show that rounder shapes have lower specific resistance and are thus more energy efficient. The highest velocity achieved in this work is 26 module lengths per second (1.6m/s), which is believed to be the fastest gait yet implemented for an untethered modular robot.;The second locomotion principle is a novel biologically-inspired legged style of locomotion. Passively compliant leg attachments are utilized to achieve a dynamic running gait using body articulation. We used gradient search to optimize the running gait parameters on two sets of legs with different stiffness. With experimental data and analysis we show that the softer legs run like a Lateral Leg Spring (LLS) model.