Why does a moving bicycle stay upright?

A bicycle stays upright when rolling forward because it is steered in the direction of any lean that develops. Turning into the lean moves the ground support back under the center of mass. A rider normally provides the steering, but many bikes at the right speed can produce the needed steering torque on their own. [1]

MIT engineer David Gordon Wilson compares this to balancing a broomstick on your palm: you keep your hand under the direction of the fall. On a bike, he says, you steer under the fall. [2]

A University of Cambridge opinion piece describes riding as a stream of tiny, mostly subconscious corrections. If you fall to the right, you steer slightly right so the wheels move back beneath you, then steer back again. This wobbling is normal and is more visible in beginners. The same piece argues that the gyroscopic effect cannot be the whole story, because a novice could then simply push off and be carried along, whereas riding has to be learned. [3]

Researchers know enough about how riders create this steering torque to build bicycle-riding robots. They are still working out exactly what riders sense (lean angle, lean rate, steering angle, or a mix), through which senses, and how they respond. [1]

Riderless bicycles can stay up thanks to self-stability, but only within a limited range of speeds, and not when stationary. Four effects can create torque about the steering axis: gyroscopic precession of the front wheel, ground forces acting through trail, the mass and its distribution in the front assembly and rear frame, and tire properties. On well-designed bikes these combine in the right proportions. [1] [4]

Trail is the distance between where the front wheel touches the road and where the steering axis meets the ground. Because the contact point lies behind that axis, leaning to one side makes the ground force turn the front wheel that way. Wilson notes that a regular bike's trail of an inch or two is enough to make it largely steer itself. [3] [2]

For years, the spinning front wheel (gyroscopic effect) and trail were the two main explanations. In 2011, researchers from the Netherlands and the U.S. built a bike with counter-rotating extra wheels to cancel gyroscopic effects, and a steering axis meeting the ground behind the contact point to remove trail. It still balanced. Its front assembly has a low center of mass, positioned ahead of the steering axis, so as the front falls it turns into the fall. Researcher Andy Ruina said the two effects are important but not necessary. [4] [5]

Equations based on Francis Whipple's 1899 model predict whether and at what speed a given bike is self-stable. They match the observed weave frequency of a riderless bike and predicted the self-stable experimental design. They do not cleanly assign credit to each factor, which some people misread as science being unable to explain bicycles. [1]