Jump to content

is there a wheel that allows you to turn off horizontal shut off?


Recommended Posts

I just wasted 10 minutes watching the first video watching people ride round a skate park with no falls.  Actually she was very pretty so it wasn't that wasn't all that wasted. Why not just provide a video that actually demonstrates what it is you want to show and preferably give me the time in the video it happens?

Link to comment
Share on other sites

A mems accelerometer and gyro don't provide 1 value, they provide 3. X Y and Z.  Even if the Z axis is still pointing straight down on a bowl and the Y axis is still moving forward, it can see that the X axis has changed by the gyro and that its experiencing gravitational forces by the accelerometer. With a gyroscope which can measure the rate of change regardless of acceleration, it always knows what way its pointed.

Link to comment
Share on other sites

4 hours ago, mike_bike_kite said:

Why not just provide a video that actually demonstrates what it is you want to show and preferably give me the time in the video it happens?

Were you able to realize the moments where the wheel’s pedals were not “flat in relation to the wall” (, front to back)?

Were there any moments where the pedals weren’t flat (front to back) in relation to the true horizon?

Link to comment
Share on other sites

3 hours ago, mrelwood said:

Were you able to realize the moments where the wheel’s pedals were not “flat in relation to the wall” (, front to back)?

Were there any moments where the pedals weren’t flat (front to back) in relation to the true horizon?

You sent me approx 6 hours worth of video and I'm "supposed" to find your proof somewhere in there? If you want to see an EUC where the pedals are not flat to the horizon then look in the first video I posted in this thread and watch from 3:00 to about 3:30. Can we move on now?

 

Link to comment
Share on other sites

1 hour ago, mike_bike_kite said:

Can we move on now?

I know I will. It’s rare that all other commenters at this forum and all information resources they provided are just wrong and only a single person knows how stuff really works. But I guess it happens.

Link to comment
Share on other sites

On 7/5/2020 at 10:30 AM, mike_bike_kite said:

You sent me approx 6 hours worth of video and I'm "supposed" to find your proof somewhere in there? If you want to see an EUC where the pedals are not flat to the horizon then look in the first video I posted in this thread and watch from 3:00 to about 3:30. Can we move on now?

 

Mr EUC Extreme have in several places explained he got a special programmed controller that do NOT cut off at 45deg. That is why he can do this on this wheel.

  • Like 1
Link to comment
Share on other sites

4 minutes ago, zeroSIXzero said:

Mr EUC Extreme have in several places explained he got a special programmed controller that do NOT cut off at 45deg. That is why he can do this on this wheel.

Even read what EUC extreme write himself on the same video you post (youtube): quote:

 

Valeriy Selitskiy
oh i see gr16 is a lot weakier here
 
 
 
 
Skjul 2 svar
EUC Extreme
+Valeriy Selitskiy It limited by the fact that it can not be tilted sufficiently. Therefore, it can not run full. Also, the driver is not as experienced. It will also be limited. My Gotway V Speed is the world's first prototype of the EUC which is able to be tilted more than 45 degrees.
 
 
 
 
Edited by zeroSIXzero
  • Like 1
Link to comment
Share on other sites

52 minutes ago, zeroSIXzero said:

My Gotway V Speed is the world's first prototype of the EUC which is able to be tilted more than 45 degrees.

Yes! Afair @EUC Extreme had a special firmware version from GW not cutting of like "normal wheels" at higher "side lean angles"

Link to comment
Share on other sites

and do you know why manufacturers would implement tilt detection in that way (against the horizon) as opposed to wherever gravity is felt to be (on a wall of death it would point into the wall)? I just can't see why they'd do that unless they deliberately wanted to make the system more complicated, more expensive and wanted the wheel to cut out on riders doing fast turns. Is there some advantage in doing it this way?

Edited by mike_bike_kite
Link to comment
Share on other sites

Euc's have to use both a gyro and accelerometer and have a properly tuned algorithm to always know it's orientation. Gyro to know it's orientation to earth's gravity, and the accelerometer to know in what direction it's moving. Do some clever math on top of it, and you can filter out any noise and always have a solid model to work with.

 


 

 

  • Like 2
Link to comment
Share on other sites

On 7/5/2020 at 3:08 AM, mrelwood said:

Ok, one more. Where does the wheel think is down during this video? And where do the pedals point during that time?

 

In that example the gyro information is completely steady the entire time, but the accelerometer data (if that sensor exists in this wheel) would've started pulled down (by gravity) at the start of the video, then spike out in front of the machine when you hit the wall, then relax to its downward direction at the end.  When accelerometer data is visualized on a live updating graph, it looks literally like a spring dangling from a point.

Edited by musk
Link to comment
Share on other sites

25 minutes ago, musk said:

In that example the gyro information is completely steady the entire time, but the accelerometer data (if that sensor exists in this wheel) would've started pulled down (by gravity) at the start of the bideo, then spike out in front of the machine when you hit the wall, then relax to its downward direction at the end.  When accelerometer data is visualized on a live updating graph, it looks literally like a spring dangling from a point.

I thought @mrelwood's experiment was quite good but I'll admit I wasn't wholly convinced. With my version of how things work the wheel would detect you'd tried to tilt the wheel forwards as you moved it and obviously the motor would move the wheel forward to counterbalance the wheel. When it hit the concrete (in my version of how things work) the wheel would feel that gravity was now infront of the wheel because of the inertia and so the wheel should try and accelerate to rebalance but obviously the concrete is there so it couldn't. I tried a similar experiment at home where just the top of the wheel connected to something solid and unfortunately (for me) the wheel didn't move forward to rebalance. I've been trying to come up with a better proof (one way or the other) but eventually lost the will to continue ;)

 

Link to comment
Share on other sites

22 minutes ago, mike_bike_kite said:

When it hit the concrete (in my version of how things work) the wheel would feel that gravity was now infront of the wheel because of the inertia and so the wheel should try and accelerate to rebalance but obviously the concrete is there so it couldn't.

This is the “zoom off to horizon by itself” view that I tried to showcase earlier.

EUC has just a single wheel. It can’t accelerate on its own, in any situation. It can only affect the tilt of the shell. It can’t create or modify any momentum.

 Power off the wheel and tilt it about. That is the movement the motor can do. And it’s the only movement it can do. Power it back on, and the wheel will slowly get back to level. This is done only by spinning the wheel. It doesn’t accelerate, it tilts.

 In my video example the wheel would have no problem applying a little more power to the incredibly powerful motor. It would’ve tilted the pedals to any position the wheel wishes.

 If you let an EUC go forward without a rider, it can’t brake at all by itself. It can only try to keep the pedals horizontal, or fail and tumble.

 You’re not at all the only one being confused like this. The precision, speed and power the EUC can utilize to keep the pedals level makes it feel just as stable as a vehicle that has a support point from another tire on the front to back axis. EUC is quite a technological marvel in that sense. Yet, all it can ever do is try to spin the single wheel, the only moving part it has. It can’t accelerate, brake or steer on its own.

Link to comment
Share on other sites

Sorry if someone already said this (I just skimmed the thread), but I agree that having a cutoff simply determined by the sideways tilt of the wheel is a major limitation of the current EUCs on the market. Having a tilt cutoff limits the type of riding you can do and increases the risk to the rider. That being said, I understand the reason for needing the wheel to cutoff as soon as possible when the rider falls off. (I've been off-roading a lot more with my MCM5 and I still get the "Gotway dance" when I drop the wheel even though the it has the normal 45 degree cutoff)

There are 3 ways that come to mind that manufacturers could implement to fix the issue:

1. acceleration detection: If you are making a turn on a sharp enough angle that the wheel would cutoff, there is going to be a lot of angular acceleration experienced by the wheel. Assuming that the wheel uses standard MEMS chips, the wheel should be able to detect this acceleration and assuming this stays above a certain threshold, the wheel should remain powered. Assuming this is all true, this would simply require a software update (after all the programming is done, of course).

2. weight sensing: I haven't ridden a Onewheel but as I understand it the front of the Onewheel has sensors to detect if you are standing on it. This allows you to come to a stop, lift the front or back of your front foot, and have the wheel shutoff so that it can be leaned back. EUC manufacturers could put similar sensors in future wheels that detect if the rider is still on the wheel and, assuming they still are, keep the wheel powered past the current maximum lean angle.

3. safety lanyard/ kill switch: The least elegant solution would be for future wheels to come with a (perhaps optional) safety lanyard similar to what is used for vehicles like jet skis. This is a switch that would have to be attached to the rider and the wheel. If the rider falls off the wheel, power is cut. If the rider is within a certain distance, the wheel remains powered, no matter the lean angle.

These are just my ideas as to how EUCs could be improved. What do you guys think?

Link to comment
Share on other sites

 

9 hours ago, Mark Wilson said:

1. acceleration detection: If you are making a turn on a sharp enough angle that the wheel would cutoff, there is going to be a lot of angular acceleration experienced by the wheel. Assuming that the wheel uses standard MEMS chips, the wheel should be able to detect this acceleration and assuming this stays above a certain threshold, the wheel should remain powered. Assuming this is all true, this would simply require a software update (after all the programming is done, of course).

The acceleration detection could be taken even further. If the threshold is reached with a moderate speed or slower, the wheel isn’t falling down, it’s being ridden. The threshold could then be increased for example to 60 degrees. 

Link to comment
Share on other sites

Mrelwood is absolutely correct.  The only way to make a EUC not cutoff at an extreme angle is to use historical data to calculate that the wheel is in fact "banking" in a turn. Also the geometry of that curve has to be known! And at that point you no longer rely just on relatively instantaneous data.  There is no way for the wheel to detect the curvature of the surface to compensate for the "upright" position relative to the SURFACE.  You cannot sense the curvature of a surface using an accelerometer.

How do you think bicycle bank on a turn?  Because it has two wheels, and the two wheels combined IS the surface curvature detector!

For an EUC to make this work, it requires a camera and live image processing.

  • Like 2
Link to comment
Share on other sites

1 hour ago, Cychotic said:

Mrelwood is absolutely correct.  The only way to make a EUC not cutoff at an extreme angle is to use historical data to calculate that the wheel is in fact "banking" in a turn. Also the geometry of that curve has to be known! And at that point you no longer rely just on relatively instantaneous data.  There is no way for the wheel to detect the curvature of the surface to compensate for the "upright" position relative to the SURFACE.  You cannot sense the curvature of a surface using an accelerometer.

How do you think bicycle bank on a turn?  Because it has two wheels, and the two wheels combined IS the surface curvature detector!

For an EUC to make this work, it requires a camera and live image processing.

Is there a sensor that allows you to measure the acceleration force (normal force as far as pushing the wheel into the ground along its yaw axis is concerned) and ALSO the (always pointing down) gravitational force?

Because if you have both, you could switch the wheel off only if

a) the angle between gravity and wheel is too much (e.g. 45°)

AND

b) the normal force is too weak to push the wheel into the ground

This would allow you to ride "banking" turns if you have enough momentum, right? The wheel would only switch off if it is tilted AND too slow to keep going. (Probably the cut-off angle would be dynamic, depending on the wheel tilt and yaw acceleration).

That's what I have been thinking. Sideways wheel tilt (riding a banking turn) or forward/backward wheel tilt (riding a looping*) would only switch off the wheel if it does not have enough momentum that would allow it to keep going.

If such a sensor exists, would that work or is there a conceptual problem somewhere?

[* The self-balancing of such a wheel would have to be relative to the yaw acceleration direction instead of the gravity direction. but then you could ride a looping, right? I always wondered if that would work.]

Edited by meepmeepmayer
  • Like 1
Link to comment
Share on other sites

Continue on my above post, I think you can make it work using the extra downward acceleration (relative to the wheel) while the wheel is on an angle, but my gut feeling tells me that the wheel will have severe pedal dip during the turn and would need more sensors to make it work well.   I imagine it will have so many corner cases where it won't work.  You almost have to preprocess the path inorder to properly compensate smoothly.  A person on a mechanical unicycle can do this obviously because he can visually see the curve and are anticipating it.  You simply cannot do that blind folded on an unknown curve.

  • Like 2
Link to comment
Share on other sites

4 minutes ago, meepmeepmayer said:

Is there a sensor that allows you to measure the acceleration force (normal force as far as pushing the wheel into the ground along its yaw axis is concerned) and ALSO the (always pointing down) gravitational force?

There are MEMS sensors that include a magnetometer, so yes. But in my understanding the gravitational “down” can already be calculated with a reasonable precision from the accelerometer data alone, so the sensors used in EUCs could possibly already be capable of what you propose.

4 minutes ago, meepmeepmayer said:

This would allow you to ride "banking" turns if you have enough momentum, right?

As far as I can see (which may not be very far compared to some within this subject), yes. I think it should also be timed to allow for small jumps while banking, but that in turn would not engage the cut-off if the wheel falls down rapidly on level ground. So we would be back to the GW dance...

 

4 minutes ago, meepmeepmayer said:

That's what I have been thinking. Sideways wheel tilt (riding a banking turn) or forward/backward wheel tilt (riding a looping!)

It can’t work front to back though. The wheel’s #1 purpose is to keep the pedals horizontal front to back. That’s what allows us to accelerate and brake in the first place. A wheel that would follow the surface would be useless when riding up curbs or roots for example. Or just trying to accelerate up a steep hill. Or jumping off from a shallow skate pool. Etc.

 Which brings me to the issue of how far such a wheel then could be tilted to the side.

As I wrote above, keeping the pedals horizontal is something that can’t be touched. It has to be the #1 operating principle at all times. When the wheel is tilted more and more to the side, changes in the wheel’s directional heading from the rider’s perspective turn into acceleration and braking. This is what makes the MSX go wild when it tilts to the side when trolleying, a small lift from the handle turns the front of the wheel towards the ground, which results in acceleration.

 Imagine a wheel riding at the edge of a pool, tilted 90 degrees from horizon, and trying to keep the pedals horizontal at the same time. As the rider turns down towards the bottom of the pool, the pedals tilt downwards in relation to true horizon, so the wheel will accelerate until the pedals are again horizontal. It would not be a very stable feeling, I’d imagine, pedal dip in turns like no other.

Link to comment
Share on other sites

55 minutes ago, meepmeepmayer said:

Is there a sensor that allows you to measure the acceleration force (normal force as far as pushing the wheel into the ground along its yaw axis is concerned) and ALSO the (always pointing down) gravitational force?

Because if you have both, you could switch the wheel off only if

a) the angle between gravity and wheel is too much (e.g. 45°)

AND

b) the normal force is too weak to push the wheel into the ground

This would allow you to ride "banking" turns if you have enough momentum, right? The wheel would only switch off if it is tilted AND too slow to keep going. (Probably the cut-off angle would be dynamic, depending on the wheel tilt and yaw acceleration).

That's what I have been thinking. Sideways wheel tilt (riding a banking turn) or forward/backward wheel tilt (riding a looping*) would only switch off the wheel if it does not have enough momentum that would allow it to keep going.

If such a sensor exists, would that work or is there a conceptual problem somewhere?

[* The self-balancing of such a wheel would have to be relative to the yaw acceleration direction instead of the gravity direction. but then you could ride a looping, right? I always wondered if that would work.]

Yea, I think you could put sensors in the pedal or axel to measure the force applied.  That's definitely a possible solution.  But what we currently have in our wheel is not sufficient.  It will only work in a very limited scenarios.

  • Like 1
Link to comment
Share on other sites

11 minutes ago, mrelwood said:

It can’t work front to back though. The wheel’s #1 purpose is to keep the pedals horizontal front to back.

"Horizontal" compared to what?;) This is my idea about that:

1 hour ago, meepmeepmayer said:

[* The self-balancing of such a wheel would have to be relative to the yaw acceleration direction instead of the gravity direction. but then you could ride a looping, right? I always wondered if that would work.]

Plausible?

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...