[Active] Bicycle Seatpost System by Shimano

In this article, we’ll be discussing a Bicycle Seatpost System by Shimano, US patent 11066118. The publication date is July 20^th, 2021 and the filing date is Jan 26^th, 2018. This has been granted.

Shimano are showing a road/gravel-type bike, but this technology can probably be used on any bike. Why would they limit themselves?

Brief Summary (tl;dr)

Shimano are working on yet another active component. This time, it’s the dropper. Like always, the bike has some sensors on it. In this case, the bike can sense speed, roll (leaning), and gravitational acceleration. The dropper itself is a screw-type linear actuator, where there’s a screw attached to a nut and the nut is attached to the upper tube. The screw is rotated by a motor. Depending on the direction of the motor, the seat will move up and down without the rider applying any pressure to the seat, just a button.

So, there’s two scenarios where the dropper moves, automatically. First, the bike will sense leaning vs velocity and satisfy an inequality that compares centripetal vs centrifugal forces, attempting to keep centrifugal force lower than the centripetal forces (check out the article for a deeper dive into this). So, if the rider is going too fast for a turn, the seat lowers to lower the rider’s CG, which then in turn reduces the centrifugal force at the tire so the bike can ‘maintain stability’.

The next scenario is some nut-job doing jumps on a fucking road bike. The system can determine gravitational acceleration during a jumping event. The system detects the bike going off the dank jump, determines zero gravitational acceleration when the bike is in the air, and detects a hard-ass landing. The seat would go down, automatically, in during the start of the jump, through the landing, and the seat would be up during normal riding.

Background

Droppers. Not sure if I need to explain the background on this. Push a button, they go up. Push the button again, and they go down.

Everyone and their mother are making droppers, except Shimano. Hell, SRAM are even on electronic and wireless droppers — time for Shimano to make one. They don’t make one, right?

Why

As always, Shimano never have a problem statement. Oftentimes, there isn’t even a problem – they just do stuff.

Generally, the present disclosure is directed to various features of a bicycle seatpost system that change a height of a height adjustable seatpost in accordance with detected information about a change of orientation (e.g., a cornering orientation or a jumping orientation) of a bicycle relative to a ground.

How about that for some legalese. They’re just saying the seatpost changes position based on detecting information.

What

Alright, lets get through the boring shit. FIG. 2 shows the entire system in diagram-form. Shimano are showing a lateral inclination sensor S1 and a vertical acceleration sensor S2. The lateral inclination sensor will sense how far the bike is leaning. Think MotoGP-style roll. It’ll return an inclination in degrees. The vertical acceleration sensor will sense a change in velocity in the vertical direction (duh). Think jumping, climbing, bumps, or chatter.

We also have a speed sensor S3 measuring speed and a crank angle sensor S4 measuring… crank angle.

Shimano say the lateral inclination sensor and the vertical acceleration sensor can be a cycling computer (Garmin, Wahoo, etc.), but it can also be your phone:

Typically, a “smart” mobile cellular telephone is provided with an accelerometer and a gyroscope. In the case of “smart” mobile cellular telephone as the mobile device 32, the lateral inclination sensor S1 and the vertical acceleration sensor S2 can be formed by the accelerometer and gyro- 35 scope of the mobile device 32.

The information from all those sensors is sent to the seatpost, which includes a controller. The controller then tells the actuator (motor) what to do based on the sensed information. Remember, active shit = sensors + controller + actuator.

The electronic controller is configured to control the electric actuator to change a height of the height adjustable seatpost in accordance with detected information about a change of orientation of a bicycle relative to a ground.

FIG. 3 shows this new dropper. It’s very clear that there are missing component details with this post, so we can assume that this is just a generalized image about the overall concept.

FIGs. 4 and 5 show the actuation of the dropper. This uses a screw-type linear actuator to move the seatpost up and down. We’ve already seen this idea from Trek. It’s a super simple idea. The screw 26 is attached to a motor 16 at one end (bottom) and to a nut 28 at the other end (top). The nut is fixed to the upper tube of the dropper. The motor will rotate the screw inside the nut, and as a result, the upper tube will go up and down, depending on which direction the screw is rotated.

Now we’re getting into the meat of this one. First, Shimano show the equation to determine operational inclination angle θ, where m = mass, g = gravity (9.81 m/s), r = curve radius, and v = velocity.

Shimano also has an example scenario. If the bike is traveling 10kph on a 5-meter curve radius, the approximate inclination angle θ is 9°. So, why are we calculating this? The system is going to use this value (degree) as a threshold value (reference angle α ) shown in FIG. 6, so a theoretical maximum angle, and compare it to an inclination angle θ, so the actual angle of the bike.

Ideally, the system will compensate for the coefficient of friction of the road for stability. In other words, the reactionary force (opposite force) of the horizontal centripetal force is determined, as the rider goes around a corner. In even more detail, horizontal centrifugal force (mv^2/r) should be below the forces of the horizontal centripetal force (μmg) to ‘maintain the stability of the bicycle’.

So, wtf did I just say? In short, the system is trying to keep the bike stable, by keeping the centrifugal force below the centripetal force. For those that forgot (like me), centrifugal force is the force felt by an object in a curved path, so it’s the outward vector force. Alternatively, the centripetal force is the opposite, which is the force needed to keep an object on the curve. Both are perpendicular to the turning arc. Centrifugal is perpendicular pointing outward; centripetal is perpendicular and pointing inward. In an ideal scenario, centrifugal and centripetal are equal and opposite, meaning there is no slip, and the bike is going in it’s intended direction.

To maintain the stability of the bike, the system solves for the inequality:

where  = coefficient of friction. As you can see, the system wants the centrifugal force lower than the centripetal force. So, how does it do that? Pretty simple, the system lowers the seat post, which lowers the rider, which lowers the centrifugal force, which maintains stability.

So, here’s the basic idea, based on all the shit I just told you. The reference angle α is an angle of lean determined by the user (or the computer, see next paragraph). The inclination angle θ is the actual angle the bike is at, at any given time. FIG. 6 shows reference angle α and FIG. 7 shows inclination angle θ. In short, the system will raise and lower the seat post, automatically, to adjust the center of gravity (CG) to satisfy the inequality we just discussed, to keep the bike stable based on the velocity of the bike. It wants the inclination angle θ to be at, or close to, the reference angle α. That’s it.

‘…it is possible to change a height of a seatpost to an appropriate height for traveling a corner.’

FIGs. 6 and 7 show a general scenario. Hey Shimano, where the fuck is this guy’s helmet? I put one on there because I’m not endorsing such reckless behavior. Wear your helmet, kids.

If the bike is going too fast for the curve, the seat lowers. If the bike is going too slow for the curve, the seat raises.

Here’s a cool part. The user can actually select the reference angle themselves. That means the reference angle (theoretically maximum angle) the bike can lean is determined by the rider. This is interesting because the rider can set that shit hella low so they can try to get a knee down, Marquez-style. But, Shimano also say the system can set the angle based on weight, height, etc.

It’s very, very important that the phone, Garmin, etc. are mounted perfectly dead center of the bike so the system can calculate an accurate angle.

‘To facilitate the detection of the inclination angle θ, the mobile device is mounted on the handlebar stem so that the lateral inclination sensor S1 and the vertical acceleration sensor S2 lie substantially on the center longitudinal plane (i.e., the reference upright plane RP) of the bicycle.’

FIGs. 8 and 9 show the same concept, but with the rider leaning outside the center axis (reference plane RP), which would probably be more realistic. This system can compensate for a change in position and use the reference upright plane RP’ (RP prime) in the calculation. As you can see, the reference upright plane RP’ travels from the contact point CP through the determined center of gravity CG. CG is bike + rider, not just the rider.

Alright, this idea goes even further than just leaning. If anyone has ever wondered how companies like Garmin determine ‘jumping distance/time’ and shit like that, this is going to be very similar to how they do it. FIG. 10 shows an example jumping scenario.

Here, we see a graph showing a gravitational acceleration value plotted over time. First, I have to say this is an annoying graph, as it’s showing an absolute value of gravity, so they’re removing the negative. Gravity should be negative, based on any normal coordinate system (Csys), but technically neither is wrong. A gravitational sensor like this will detect current acceleration and determine the difference between gravity and current acceleration. In this case, the rider isn’t going up or down, so the difference is (9.81 m/s – 0 m/s) = 9.81 m/s

In section 1, we have no jumping, so just riding along. Notice how the gravitational acceleration is pegged around 10 m/s. As the rider begins a jump (in section 2) the detected gravitational forces increase, because this dude is going off a fucking jump on a road bike…

Then, in section 3, he’s doing a sick-ass jump, Rogatkin-style. Notice there is zero detection gravitational acceleration. Why do you think that is? It’s because everything, always, accelerates at -9.81 m/s toward the ground, so the system will detect zero acceleration because the bike is moving at the speed of gravity. Same calculation as before, the bike is going 9.81 m/s, so it’s (9.81 m/s – 9.81 m/s) = 0 m/s.

In section 4, this dude lands his sick-ass jump and the system detects the bike and rider reacting to the ground, so the data is super noisy. Lastly, section 5 is homeboy riding off to fame, because no one jumps a road bike.

So, why am I telling you about sick-ass jumps? Well, not only does this system adjust the seat post based on cornering and cornering speed; it also adjusts the seat post for jumping. I’m sure you can assume what I’m about to say. When the system detects the start of a jump, the seat lowers as to not impede the rider from doing a knack-knack. The seat can stay lowered during airtime, then raise back up when the detected gravitational acceleration normalizes near 9.81 m/s.

Conclusion

Speaking about the screw-type linear-actuator seatpost (and going off on a tangent), there’s a concept in auto racing (any racing, honestly) where a convergence of ideas typically suggests a correct concept. For example, the current turbo-era we’re living in lends itself to the power vs. fuel economy battle, and ever stringent EPA standards. The EPA says “You need to meet these standards by this date” and the manufacturers need to figure out a way to do it. Turbos recycle typically-wasted energy and transfer that energy back to the motor. Currently, nearly every major manufacturer has their own version of a small 4-cylinder turbo motor, as opposed to V6’s and V8’s from 20 years ago. In short, they’ve figured out that they can use a smaller motor with a turbo and get the same power using less fuel. This is a convergence of ideas and has proven to be the most correct answer, for now. Do you see a new convergence of ideas coming through the automotive industry, with respect to this same concept?

I bring this up because we’re seeing a convergence of ideas relating to a linear-actuated dropper. Both Shimano and Trek (probably others too) have developed the same general idea without input from each other (conjecture).

Since we’ve seen the screw-type linear actuator dropper before, I’ve had a lot of time to think about it (along with lots of input from others). I believe this will be the standard design for droppers in the near future. There are numerous benefits to this design. First, it’s a simpler design. Yes, there are subtleties that will need to be addressed just like any new technology, but the overall design is extremely simple and is already implemented in a shit-load of other applications (3D printing tables, for example). Just a screw, bolt, and a few bearings, and you’re good to go.

Second, it allows for movement in two directions without direct contact with the seat. Homeboy Robert Johnson commented on Jan. 6^th with this. As you know, conventional droppers must be sat on to get back into the lowered position, so there must be an external force to move them. With the linear actuator, there is no need for an external force. It’ll just move with the flick of a switch.

Third, as we’ve seen here, this now allows for automatic and active actuation in both directions. Conventional designs can’t do this.

I’ll admit, this was a complicated one. God, I hope it made sense… Please let me know if you see an error.

Shimano out in 3000.