we have yet devised.

But perhaps even more incredible than humans riding bicycles is the fact that bicycles

can ride themselves.

Yes, once they’re set in motion at a sufficient speed, bicycles can stay upright without any

human intervention.

A common misconception is that bikes stay up because of conservation of angular momentum

– that is, since the wheels are spinning, if the bike tips to one side there’ll be

some sort of countering force from the wheels that keeps the bike upright.

But there’s an easy way to see this explanation is wrong: simply lock the handlebars in place

and a moving bike will fall over just as easily as a stationary one.

Another common misconception is that bikes stay upright because of their forward momentum.

However, if you knock a ghost-riding bicycle sideways, it’ll change directions and then

continue merrily on its way – plainly changing its momentum, but nevertheless staying upright.

What we do know about how conventional bikes stay upright on their own is this: when a

moving bike starts leaning to one side, it also automatically steers towards that side

a little bit.

The result is that the wheels come back underneath the center of mass, keeping the bike balanced.

And there are three main mechanisms responsible:

First, because of the backwards tilt of a bike’s steering axis, its front wheel actually

touches the ground slightly behind that axis.

This means that when the bike leans to the left, the upward force from the ground acts

to turn the wheel and handlebars to the left, helping the bike steer its wheels back underneath

its center of mass.

Second, the weight of a bike’s front wheel and handlebars is generally distributed slightly

in front of the steering axis, so when the bike leans to the left, the downward pull

of this mass also helps turn the front wheel to the left, the same way divining rods will

turn towards whatever direction you tilt your hands.

Third, there is indeed a gyroscopic effect from the wheels, but it doesn’t keep the

bike upright on its own.

Instead, it helps steer: as Destin and Carl demonstrate excellently in this video about

how helicopters work, trying to tilt a spinning object makes the object tilt as if you pushed

it at a point 90° away from where you did – it seems spooky, but basically the effect

of your torque lags behind where you push.

Now imagine this happening vertically on a bike, and you can see that the gyroscopic

precession from the bike’s leftward lean makes the front wheel turn to the left, again

helping steer its wheels back underneath its center of mass.

In short, a normal bicycle is stable thanks to a combination of the front wheel touching

the ground behind a backwards-tilted steering axis, the center of mass of the front wheel

and handlebars being located in front of the steering axis, and the gyroscopic precession

of the front wheel, all of which help the bike automatically steer its wheels back underneath

it when it leans.

At least, when it’s moving forwards at the correct speed.

If the bike’s going too slow, it won’t turn quickly enough to keep from crashing

into the ground.

And if you push the same bike backwards, the gyro effect will reverse but the other two

effects won’t, with the result that the wheels are steered out from under the bike

when it leans.

What’s more, none of these three mechanisms is, on its own, the secret to bike stability:

here’s a bicycle that has no gyroscopic effect and whose front wheel touches the ground

in FRONT of the steering axis yet which is stable without a rider.

Here’s a stable rear-steering bike, and here’s a design for a stable bike where

the steering axis tilts forward instead of back.

On the other hand, I made my own bike totally unstable just by adding some extra weight

behind the front fork.

There are clearly a lot of different variables that can be combined in various and surprising

ways to make stable and unstable bicycles.

Adding a human to help with steering and balance can sometimes make unstable bikes stable,

and I imagine a rider would also make some stable bikes unstable.

But amazingly, even for a riderless bike, science currently doesn’t know what it IS

about the special combinations of variables that enables a bike to stay up on its own.

We just know that some combinations work, and others don’t.