Decoding Stride Variability in a Galloping Sled Dog Team

Harnessed into a synthetic network of tethers, sled dogs form an artificial collective where cohesion is enforced and arrangement is predetermined. Using custom animal-borne sensors, we collected multi-sensor movement data from a team of sled dogs to investigate their collective locomotion. While maintaining a steady pace, the sprinting sled dogs predominantly galloped, yet exhibited a tendency to alternate between distinct galloping patterns—specifically rotary and transverse—as well as vary their leading limb. Using the collected acceleration data, we performed unsupervised clustering of the trajectories to identify behaviorally distinct stride types and the transitions between them. We then ask whether the dogs are choosing from a limited set of accessible gait patterns to maximize their stability amid the perturbations induced by their harnesses and lines—a potential manifestation of quadrupedal multi-stability.