Walking is like slithering: A unifying, data-driven view of locomotion

What is it about?

Animals move by generating forces against their surrounding environment. We present a unifying framework that applies to slithering, walking, and swimming in viscous (Stokes) fluids. Our results show a surprising fact: both with and without slipping, these motions can be modeled without knowing the forces, because these motions are "principally kinematic". This was previously known for slithering and Stokesian swimming, and we have discovered it also applies to multi-legged walking animals (ants) and robots, both with and without slipping. We describe procedures for building principally kinematic models from observational data, and for testing their validity. Our work provides a universal model for locomotion that applies whenever the movement is dominated by continuous friction with the environment. This model can be applied in robot design and motion planning, and provide insight into the evolution and control of legged locomotion.

Featured Image

Why is it important?

Animals may use this kind of modeling when learning how to move and when planning motion, and this may provide insights into the evolution of motor control and locomotion. Principally kinematic models may enable roboticists to use multi-legged robots that, despite their great promise, have seemed too difficult to model and control.