Duty Factor Predicts Robust Constrained Quadrupedal Locomotion Across Gait Types

TL;DR

Duty factor, the fraction of a stride during which a leg is in contact with the ground, is a more reliable predictor of locomotion robustness in quadrupedal robots than gait type. The study demonstrates that duty factor consistently influences error convergence across multiple control methods, including optimization-based, learned, and model predictive control. Experiments on a physical robot confirm that adjusting duty factor improves stability on narrow terrain.

Why It Matters

Robust locomotion in unstructured environments is critical for real-world deployment of legged robots. By identifying duty factor as a key control parameter, this work enables simpler, more adaptive strategies for maintaining stability without redesigning entire gait patterns. This insight can streamline control design across diverse robotic platforms.

Key Findings

• Duty factor outperforms gait type in predicting local error convergence during locomotion.
• A learned controller can use duty factor as a low-dimensional parameter to adapt to narrow-terrain challenges.
• The relationship between duty factor and robustness holds across different control frameworks and is validated in physical experiments.