Shimano is developing AI suspension adjustment
While we're still waiting eagerly for the launch of Shimano's new electronic Di2 XTR mountain bike groupset, today, we've uncovered a freshly filed patent showing what looks to be an automatic suspension and dropper post system using a 'learning model' or AI to you and me.
- Is Shimano developing an automatic suspension adjustment system?
- Is SRAM looking to kill the mech hanger?
- Is Specialized about to revive Trust's linkage fork?
The patent filed on 23 June 2020 is titled 'Control device, creation method for learning model, learning model, computer program and storage medium'. And that's the big giveaway that Shimano's creating its own AI application.
But what for? Well, suspension adjustment and dropper post control. The patent's abstract states that 'the device controls a telescopic mechanism by output information related to actuating the telescopic mechanism that is output from a learning model in associate with input information relating to travelling of the human-powered vehicle'. This suggests that Shimano's new feature uses information gathered by a telescoping component's behaviour, be that a fork, shock or seat post.
It looks like Shimano's new technology is controlled or accessible through a screen, whether that's a device on the bike or a phone, and the images and the patent's summary suggest that it can learn a course (downhill or cross-country). It can store that information and use it to adjust the suspension and seatpost to the best settings for the rider, for that portion of the track. This can be seen in the 're-learning mode' and 'automatic control mode' in the image below.
Other images show that it may be possible to provide feedback or add some kind of manual adjustment for the system to learn better.
Does this have a use in modern mountain biking?
Many will be reaching for their pitchforks but this kind of technology can open up a whole new realm of suspension adjustment. But it may only appeal to those racing at the very top of their disciplines.
This Shimano system seems like it will be able to learn a track and automatically adjust the best-performing suspension setting and dropper post position for particular sections of a given course. And that's through rider feedback and likely multiple laps of the same track. So for top-level competition, if successful, we could speculate that this could potentially take the work away from on-site suspension mechanics. But let's face it, that's a long shot.
Throughout last year, we saw riders such as Pauline Ferrand-Prevot and Loic Bruni hiding some kind of electronic lockout tech on their race bikes, so methods of electronically locking out suspension already exist.
And of course, Fox Live Valve and RockShox Flight Attendant have been on the market and the race scene for years now. Both of these can be seen as similar to what Shimano is proposing but neither can learn a whole race track. We bet this tech won't be picked up by the enduro crowd...
The patent goes on to say that 'proper skill is required for the rider to appropriately control the seatpost and suspension during riding'. So it's clear that Shimano is looking to take the brain power away from the rider and create a more advanced and possibly easier-to-use automatic suspension adjustment.
2 comments
Hang on until the system can actually see what the rider sees in front of them (and more) then the ai is similar ly going to be guessing where your dropper post will be.
Great until the learning goes awry and fires the dropper into your nads. Perhaps it's an early example of an AI extinction level event