Why robots move better when they stop trying so hard

Efficient robotic movement often comes from working with physics, not against it. This work formalizes "natural locomotion"—motion where internal oscillators exchange energy with the environment in a balanced way, using zero net external power over each cycle. The authors develop a mathematical framework (Natural Locomotion Manifold) that identifies which passive mechanical designs can sustain such cycles, tested on systems like pendulum-driven cars. The approach shifts robotics design from trajectory tracking to finding architectures that inherently want to move.