Pre-programmed fixed gait patterns are used to achieve effective robotic locomotion. However, this approach limits the ability to perform general locomotion in diverse terrains and at different speeds.
A recent paper shows that minimizing energy consumption plays a key role in the emergence of natural locomotion patterns in flat and complex terrains at different speeds.
Legged robots are being developed by different research teams globally. Image credit: Boston Dynamics
The researchers use energetics to design an end-to-end learning framework and display the resulting gait patterns in a real quadruped robot. The model-free reinforcement learning is employed as the optimizer to make the quadruped move forward while simultaneously minimizing the bio-energetics constraints.
Additionally, a distillation-based learning pipeline is used to achieve smooth gait transition as the target speed is changed. It is shown that the emergent gaits correspond to conventional animals without any pre-programming.
Legged locomotion is commonly studied and expressed as a discrete set of gait patterns, like walk, trot, gallop, which are usually treated as given and pre-programmed in legged robots for efficient locomotion at different speeds. However, fixing a set of pre-programmed gaits limits the generality of locomotion. Recent animal motor studies show that these conventional gaits are only prevalent in ideal flat terrain conditions while real-world locomotion is unstructured and more like bouts of intermittent steps. What principles could lead to both structured and unstructured patterns across mammals and how to synthesize them in robots? In this work, we take an analysis-by-synthesis approach and learn to move by minimizing mechanical energy. We demonstrate that learning to minimize energy consumption plays a key role in the emergence of natural locomotion gaits at different speeds in real quadruped robots. The emergent gaits are structured in ideal terrains and look similar to that of horses and sheep. The same approach leads to unstructured gaits in rough terrains which is consistent with the findings in animal motor control. We validate our hypothesis in both simulation and real hardware across natural terrains. Videos at this https URL
Research paper: Fu, Z., Kumar, A., Malik, J., and Pathak, D., “Minimizing Energy Consumption Leads to the Emergence of Gaits in Legged Robots”, 2021. Link: https://arxiv.org/abs/2111.01674