Category Archives: robotics

Self-balancing motorcycle

Road and Track magazine has reported on a self-balancing motorcycle from Honda.

How does the motorcycle balance itself? As shown in videos with the Road and Track article, the front fork is hinged so the front wheel can move forward, increasing trail — the distance of tire contact patch behind the steering axis. Then automated steering motions shift the mass of the machine and rider slightly side to side to maintain balance. The machine is also reported to be able to perform a conventional track stand, like those done by bicyclists — turning the front wheel at an angle to one side, and maintaining balance by propelling the machine forward and backward.

These capabilities require that the motorcycle have special control mechanisms, a computer on board to operate them, and sensors to report the machine’s orientation — which can be tricky on a single-track vehicle, because it can balance even when leaning into a turn. The Honda feature is described as for low speeds and when the motorcycle is stopped. That would avoid the issue with leaning. An internal-combustion engine would require a special and complicated transmission to drive the motorcycle subtly backward and forward on short notice.  In any case, electrical motors are needed to adjust the fork angle and make subtle steering corrections. An all-electric motorcycle is simpler and can be used indoors, as shown in the videos in Road and Track. They show no exhaust pipe.
In a discussion on Facebook, Jim Lindner wrote:

Extending the wheel base and fork angle gives the ability to move the mass of the engine right and left, a form of weight shifting they likely did not count. A motorcycle has a bit more mass than a bicycle improving the resistance to change, but with a little leverage or sudden shift of load I think this system’s ability to correct will easily be overloaded.

This brings up the concept of the  operational design domain: the range of conditions under which robotic features work. Automated-vehicle researchers and engineers use this concept. ODDs range from cruise control and anti-lock braking up through robotic crash avoidance for a vehicle otherwise under the control of a human driver, and onward to driverless operation under increasingly more challenging conditions.

The ODD for an automated two-wheeled vehicle does not include slippery surfaces or steering into a curb at a low angle unless it has sensors which allow it to avoid these hazards.  To avoid falling over, it must avoid these conditions as a skillful human driver would. This capability is far in the future.