Shimano MTB AI suspension control to take guesswork out of ride setup and adapt to rider and course requirements

As batteries have become smaller, and electronics more efficient, they have started creeping into sports like mountain biking. Never one to miss a hype train, the MTB industry now seems to be hopping onto the AI and machine-learning bandwagon, with Shimano recently filing a patent application for AI-powered suspension and dropper-post control.

The gist of the patent [PDF] is that Shimano wants to use machine learning to adjust suspension settings — like compression, rebound, and damping — and seat post height on the fly, depending on where and how riders are riding. According to the patent, Shimano's proposed system will make use of an array of sensors to detect everything from speed, torque, and acceleration to cadence, incline or angle, and pneumatic pressure (presumably of the suspension) to set parameters for the neural network.

The neural network processes this wealth of information alongside rider inputs and the resulting output to learn the rider's preferences and adjust suspension and seat post height accordingly. Shimano explicitly details a method to retrain the system in the case of undesirable results, with user feedback being used to guide the system in its training.

The automatic control modes will be course-dependent, and users can select the desired course using a digital cockpit. Shimano's stated goal with the AI-powered suspension tuning is to increase rider comfort and overall efficiency of the system, making it particularly suited to racing, where every watt counts.

While the bicycle pictured in Shimano's patent application is a drop-bar mountain bike, which would generally be used for less hardcore trails and courses where efficiency matters more, it's not difficult to imagine the same system working for something like a trail bike, or even more aggressive enduro, and downhill bikes.

The human-powered control aspect of the system presents something of a problem when it comes to optimisation. Particularly inexperienced riders may not know what's best for them. While bike setup is usually handled by experienced race mechanics with feedback from the rider, a mechanic can push back against a rider's request or dig deeper into why a rider prefers one setting over another, suggesting an alternative instead. This AI-powered system is completely at the whim of the rider, which might cause riders to stick to bad habits instead of looking outside their comfort zone.

For the time being, this is also just a patent, and there's no guarantee that Shimano will ever produce its AI suspension system at all. However, it would not be surprising to see more of the cycling industry flock to AI trends, like the ChatGPT-enabled Urtropia e-bike from last year.

Buy a Peak Design Out Front bicycle and motorcycle handlebar phone mount or a TranzX Jump Seat Cable-less Dropper Seat Post on Amazon.