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

[meepmeepmayer]

6 minutes ago, Cychotic said:

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.

Just a sensor good enough to precisely measure gravitational and yaw accelerations independently. If that is possible...

[Cychotic]

You would achieve that if we can also measure how much force is applied to the pedals or axle. Combine that with existing accelerometer, it is definitely a step forward.  I hope I didn't misunderstood your idea.

Edited July 13, 2020 by Cychotic

[mrelwood]

3 hours ago, meepmeepmayer said:

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

Plausible?

I was referring to true horizon.

 As I tried to explain, there’s no way an EUC could balance front to back to the riding surface. The pedal angle would go crazy on every root, bump and curb. True horizon is the only possibility.

[meepmeepmayer]

1 hour ago, mrelwood said:

 As I tried to explain, there’s no way an EUC could balance front to back to the riding surface. The pedal angle would go crazy on every root, bump and curb. True horizon is the only possibility.

It wouldn't balance relative to the riding surface (maybe in effect, but not by design). That's not what I meant.

It would self-balance relative the the plane perpendicular to the direction of the strongest "downward" acceleration (not real downward, downward as in the rider is pushed into the wheel). Usually that direction would be gravity and it would balance relative to the true horizon, but in banking turns or loopings that plane could vastly vary. E.g. if you ride fast through a banking turn, the plane would temporarily be (almost) parallel to the surface you are riding on, and in a looping the plane would do a 360° rotation.

You just need to be fast enough (enough momentum) so you are pressed onto the wheel for this to work.

So if you used that (constantly varying) plane as reference, the wheel could behave like any other vehicle that is not dependent on how the true horizon. You can do loopings, you can ride banking curves without the wheel switching off due too much sideways tilt (because from it's perspective, there is almost no sideways tilt).

It would behave like any EUC except when gravity is no longer the dominating force. That can be in banking turns/loopings. And I think it would have to tilt the pedals if you accelerate forwards (tiltback) or brake (tiltforward). Basically the pedals would try to "fight" you or indirectly follow the terrain contour. Which may actually be good, you could exert more power on the wheel this way?

Does that make sense?

TLDR: The net direction that pushes you against the wheel is used to define what (constantly varying) horizon the self-balancing uses. Usually that's gravity pointing into the ground, resulting in the true horizon, but not always.

-

It will probably indeed go crazy from bumps, though, so I'm not sure if this is practical. But I wonder if in theory this can work.

[Mark Wilson]

8 minutes ago, meepmeepmayer said:

It wouldn't balance relative to the riding surface (maybe in effect, but not by design). That's not what I meant.

It would self-balance relative the the plane perpendicular to the direction of the strongest "downward" acceleration (not real downward, downward as in the rider is pushed into the wheel). Usually that direction would be gravity and it would balance relative to the true horizon, but in banking turns or loopings that plane could vastly vary. E.g. if you ride fast through a banking turn, the plane would temporarily be (almost) parallel to the surface you are riding on, and in a looping the plane would do a 360° rotation.

You just need to be fast enough (enough momentum) so you are pressed onto the wheel for this to work.

So if you used that (constantly varying) plane as reference, the wheel could behave like any other vehicle that is not dependent on how the true horizon. You can do loopings, you can ride banking curves without the wheel switching off due too much sideways tilt (because from it's perspective, there is almost no sideways tilt).

It would behave like any EUC except when gravity is no longer the dominating force. That can be in banking turns/loopings. And I think it would have to tilt the pedals if you accelerate forwards (tiltback) or brake (tiltforward). Basically the pedals would try to "fight" you or indirectly follow the terrain contour. Which may actually be good, you could exert more power on the wheel this way?

Does that make sense?

TLDR: The net direction that pushes you against the wheel is used to define what (constantly varying) horizon the self-balancing uses. Usually that's gravity pointing into the ground, resulting in the true horizon, but not always.

-

It will probably indeed go crazy from bumps, though, so I'm not sure if this is practical. But I wonder if in theory this can work.

That seems (in my head at least) that it would work. At least, as you pointed out, on smooth surfaces like a skate park. Although going over bumpy terrain would make the "reference plane" as you called it, constantly change. I think that would be similar to rapidly switching between pressing the pedals forward and backwards on our current wheels. Would that be controllable though, I don't really know. Maybe if you add a filter it would be (though this would create more pedal dip).

 

@mrelwood I didn't think of what would happen if our current wheels reached a 90 degree (sideways) angle. You might be right, it might make the wheel to unstable. And what would happen if you went passed that, making the wheel upside down? Would the controls be reversed?

[Cychotic]

19 hours ago, mrelwood said:

I was referring to true horizon.

 As I tried to explain, there’s no way an EUC could balance front to back to the riding surface. The pedal angle would go crazy on every root, bump and curb. True horizon is the only possibility.

With current technology, that's pretty much true in my opinion.  If you design more sensors in the pedals, you can use that info to aid balancing against the surface by detecting rider's position and what type of force applied.  You would need both downward force and also torque of the riders foot. Is the riding applying force on its toes or heels and how much?  How precise will the sensor be?  How fast does it have to respond? It is still a very challenging problem.

 

Response time would be the most challenging aspect of making it work.  In general the wheel is trying to filter out noise (bumps) which slows down the response time to achieve smooth ride. And in the case of banking a surface, it's the opposite, you have to react to the curvature as fast as possible  So we can't do both at the same time.

 

Perhaps create a hybrid control approach where rider can use an handheld controller to tell the EUC to spin faster, slower, or hold constant?  This would be consist of a trigger to activate the override and lock in the current speed of the motor, then a joystick or something to increase or decrease speed.  This would take tremendous amount of human skills to master, but it's possible with training.  It would be the easiest to implement.

 

And also as I mentioned earlier cameras and imagine processing.  Essentially, if you have all the technologies required to make a semi-autonomous EUC where it has to understand it's path, it is definitely possible. Any multi-millionaires or billionaires want to fund the project?

 

 

 

Edited July 14, 2020 by Cychotic