A gait is the pattern and timing of how a legged robot moves its feet — walk, trot, gallop — and choosing and generating the right gait is the foundation of getting a legged robot to move at all.
A gait is the rhythm of walking — which feet move when. Animals switch gaits (walk, trot, gallop) depending on speed, and legged robots do the same: a slow careful gait for stability, a fast bouncy one for speed.
Before a legged robot can go anywhere, it has to answer a basic question: which foot moves when? That pattern — the rhythm of stepping — is the gait, and it's where legged locomotion begins.
What it is
A gait is the cyclic timing of each leg's two phases: swing (foot in the air, moving forward) and stance (foot on the ground, pushing). Coordinating those phases across all legs defines the gait. A slow gait keeps most feet down for stability; a fast gait has more feet in the air at once for speed.
Legs cycle between swing and stance
Every leg repeats this cycle; the gait is how the cycles of different legs are timed relative to each other.
Static vs dynamic gaits
The key spectrum:
Statically stable gaits — the robot's center of mass stays over its planted feet at all times, so it could freeze mid-step and not fall. Slow but safe (a careful crawl for a quadruped or hexapod).
Dynamically stable gaits — the robot is momentarily "falling" and catching itself each step, like a trot, run, or gallop. Faster and more efficient, but they demand active balance (ZMP, capture-point) and can't just stop mid-stride.
Animals switch gaits with speed for efficiency; advanced robots do too.
How robots generate gaits
Scripted/parametric gaits — predefined timing patterns (a fixed trot), simple and reliable on flat ground.
Central pattern generators — biologically-inspired oscillators that produce rhythmic gaits and adapt them.
Learned gaits — reinforcement learning discovers robust, natural gaits (and gait transitions) that handle rough terrain, now standard on modern quadrupeds.
Why it matters
The gait is the organizing principle of legged movement — the difference between a pile of legs and a robot that walks. Choosing the right gait (stability vs speed) and generating it robustly is the core problem of legged locomotion, from insect-like hexapods to running humanoids.
