Meeting Abstract
114.1 Monday, Jan. 7 Don’t break a leg: injury prevention, robustness and stability of legged locomotion BIRN-JEFFERY, AV*; HUBICKI, C; BLUM, Y; HURST, J; DALEY, MA; Royal Veterinary College, UK; OSU, Oregon; Royal Veterinary College, UK; OSU, Oregon; Royal Veterinary College, UK abirnjeffery@rvc.ac.uk
In uneven terrain, legged animals must avoid falling and exceeding tissue safety factors to prevent injury. Simple models of locomotion highlight a potential trade-off between stability and injury avoidance as some control strategies for stability require large leg forces. To investigate scaling effects on leg control for stability, we studied obstacle negotiation in ground birds spanning a 50-fold body mass range. We expected larger, straight-legged animals to prioritise reduction of peak forces, leading to lower robustness and stability compared to smaller species. Unexpectedly, we observed that body and leg dynamics during obstacle negotiation are similar. Furthermore, force trajectories remained similar to level terrain across species. These results suggest a common control policy, regardless of body size. We noted an asymmetry in the stance force trajectory across birds, with peak force around 30-40% of stance, which is not predicted by current models. We demonstrate that minimum-work actuation applied to an intrinsically damped leg model correctly predicts these force characteristics and suggests higher intrinsic damping in smaller species. Despite similar peak forces across terrains, the birds achieve remarkably high robustness and stability, negotiate 50% leg length obstacles with little variation in speed, and limit fall risks to only 8 in 10,000. Animals prioritise injury and fall avoidance over steady dynamics. The experimental data is consistent with a control policy involving feed-forward swing-leg control to minimise changes in passive-dynamics of stance phase. Consistent trends across a 50-fold range in body size suggest general principles which may be useful for the control of legged robots.
