Self-balancing bicycle concept

[B. Sanders]

Over the weekend, I was thinking about how schools of fish can move quickly and precisely in perfect
concert - almost moving like a single fish. I thought of cycle racing and what could be done to
instrument road bikes for safe, fast, close-quarters riding (better than human response times). I
imagined that a system could be devised to avoid/prevent/minimize crashes and subsequent pileups on
training rides, for instance.

It would be interesting to develop a bike that is self-balancing, using servo-controlled steering
correction to add dynamic balancing input. If a servo-controlled balancing mass were added in
addition to the servo-controlled steering input, you could make a pretty stable auto-balancing bike
fairly easily.

The ideal setup would allow rider steering input to act upon the servo system, which would correct
for it instantly (such as banking coordination in a sharp turn). One likely goal would be to figure
out how the rider's natural balancing rhythms could co-exist with computer controlled balancing
corrections to enhance the stability of the rider/bike/computer system. It is said that the Segway
"feels natural and stable" upon first use. Such stability and naturalness would be the goal of a
self-balancing bike system.

I suppose you could give the bike ABS braking, since you're going to the trouble of instrumenting it
with accelerometers a la Segway.

Has this been done?

Barry

 

[Jacobe Hazzard]

B. Sanders wrote:
> Over the weekend, I was thinking about how schools of fish can move quickly and precisely in
> perfect concert - almost moving like a single fish. I thought of cycle racing and what could be
> done to instrument road bikes for safe, fast, close-quarters riding (better than human response
> times). I imagined that a system could be devised to avoid/prevent/minimize crashes and subsequent
> pileups on training rides, for instance.
>
> It would be interesting to develop a bike that is self-balancing, using servo-controlled steering
> correction to add dynamic balancing input. If a servo-controlled balancing mass were added in
> addition to the servo-controlled steering input, you could make a pretty stable auto-balancing
> bike fairly easily.
>
> The ideal setup would allow rider steering input to act upon the servo system, which would correct
> for it instantly (such as banking coordination in a sharp turn). One likely goal would be to
> figure out how the rider's natural balancing rhythms could co-exist with computer controlled
> balancing corrections to enhance the stability of the rider/bike/computer system. It is said that
> the Segway "feels natural and stable" upon first use. Such stability and naturalness would be the
> goal of a self-balancing bike system.
>
> I suppose you could give the bike ABS braking, since you're going to the trouble of instrumenting
> it with accelerometers a la Segway.
>
> Has this been done?
>
> Barry

I once thought of a bike which uses gyroscopes spinning opposite to the wheels such that the bike
remained upright even during turns. This has the disadvantage of requiring that the rider 'hang
on' and not fall over the side, as well as the extra inertia of the gyroscopes for accellerating
and braking.

This is simlar to how the segway works, though.

Adam

 

[Ken]

"B. Sanders" <[email protected]> wrote in news:[email protected]:
> It would be interesting to develop a bike that is self-balancing, using servo-controlled steering
> correction to add dynamic balancing input. If a servo-controlled balancing mass were added in
> addition to the servo-controlled steering input, you could make a pretty stable auto-balancing
> bike fairly easily.

Litespeed already sells such a bike. High tech super lightweight full titanium construction, Price
is only $850

http://216.247.25.241/miva/merchant.mv? Screen=PROD&Store_Code=LS&Product_Code=T

 

[Walter Mitty]

"B. Sanders" <[email protected]> brightened my day with his incisive wit when in
news:[email protected] he conjectured that:

> Over the weekend, I was thinking about how schools of fish can move quickly and precisely in
> perfect concert - almost moving like a single fish. I thought of cycle racing and what could be
> done to instrument road bikes for safe, fast, close-quarters riding (better than human response
> times). I imagined that a system could be devised to avoid/prevent/minimize crashes and subsequent
> pileups on training rides, for instance.
>
> It would be interesting to develop a bike that is self-balancing, using servo-controlled steering
> correction to add dynamic balancing input. If a servo-controlled balancing mass were added in
> addition to the servo-controlled steering input, you could make a pretty stable auto-balancing
> bike fairly easily.
>
> The ideal setup would allow rider steering input to act upon the servo system, which would correct
> for it instantly (such as banking coordination in a sharp turn). One likely goal would be to
> figure out how the rider's natural balancing rhythms could co-exist with computer controlled
> balancing corrections to enhance the stability of the rider/bike/computer system. It is said that
> the Segway "feels natural and stable" upon first use. Such stability and naturalness would be the
> goal of a self-balancing bike system.
>
> I suppose you could give the bike ABS braking, since you're going to the trouble of instrumenting
> it with accelerometers a la Segway.
>
> Has this been done?
>
> Barry
>
>

The algorithms you are referring to are covered by the term "flocking". A google search should help.
Oh, wait : here is a URL I prepared earlier -

http://tinyurl.com/eib8

Have fun : it's an interesting topic.

--
Walter Mitty.

 

[John Tserkezis]

Jacobe Hazzard wrote:

> I once thought of a bike which uses gyroscopes spinning opposite to the wheels such that the bike
> remained upright even during turns. This has the disadvantage of requiring that the rider 'hang
> on' and not fall over the side, as well as the extra inertia of the gyroscopes for accellerating
> and braking.

Two equal mass flywheels spinning in opposite directions do not have a stabilising effect as would
one wheel.

This concep has been used in vehicles, where large mass spinning flywheels
are used to store energy that makes the vehicle move. Two are used, because if
it were only one, said vehicle would never be able to actually turn a corner.

--
Linux Registered User # 302622 <http://counter.li.org

 

[Kbh]

> > I once thought of a bike which uses gyroscopes spinning opposite to the wheels such that the
> > bike remained upright even during turns. This has
the
> > disadvantage of requiring that the rider 'hang on' and not fall over the side, as well as the
> > extra inertia of the gyroscopes for accellerating
and
> > braking.
>
> Two equal mass flywheels spinning in opposite directions do not have a stabilising effect as
> would one wheel.
>
> This concep has been used in vehicles, where large mass spinning
flywheels
> are used to store energy that makes the vehicle move. Two are used,
because if
> it were only one, said vehicle would never be able to actually turn a
corner.
>

Exactly...net rotational inetria would be zero with two opposing flywheels, the opposite of the
desired effect.

Didn't someone do a graduate thesis on the "Unrideable Bike"? I think this feature was a part of it.

 

[Ron Hardin]

John Tserkezis wrote:
> Two equal mass flywheels spinning in opposite directions do not have a stabilising effect as
> would one wheel.

A gyroscope doesn't have a stabilizing effect in the first place. It redirects force into
displacement in a direction you don't care about, but it has to be free to move in that direction.

It's mysterious because of an optical illusion: you expect force to result in displacement; but it
in fact results in acceleration, and the difference matters in the case of a gyroscope. The
displacement it experiences is in fact what is intuited except for the the appearance of the system.

You're less surprised if you bonk a pendulum bob as it hits its lowest point, and find its plane of
motion has changed at right angles to the bonk; the bob returns to the exact place you bonked it at,
that part of the plane not having moved at all. You're less surprised because you can see the bob as
a separate thing from the pendulum plane and what the bob does it intuitive.

With a gyroscope you can't see the bob; but the plane change is the same.
--
Ron Hardin [email protected]

On the internet, nobody knows you're a jerk.

 

[Jacobe Hazzard]

KBH wrote:
> Exactly...net rotational inetria would be zero with two opposing flywheels, the opposite of the
> desired effect.
>
> Didn't someone do a graduate thesis on the "Unrideable Bike"? I think this feature was a
> part of it.

I furthered the idea somewhat, what if the gyroscope had an excess of momentum, then the net
rotational inertia would be out to the left and the bike would lean *out* of turns! Then I found a
toy motorcycle that was already using the idea and lost interest..

Adam

 

[Tim Cain]

Bikes are stable and self-correcting already.

Tim.

---
Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.489 / Virus Database: 288 - Release Date: 10/06/03

 

[Tim McNamara]

In article <[email protected]>, "Tim Cain"
<tim_no1@you_know_what_to_cut_timcain.co.uk> wrote:

> Bikes are stable and self-correcting already.

If that was the case, your bike could ride without you. It would coast, riderless, without
falling over.

It's actually been done, by a British guy named Wilson who spent a lot of time trying to create
unrideable bikes (someone else referenced the URB project already) in order to find out what made
bikes rideable. It's basically all steering geometry. Bikes can be so unstable they are unrideable,
and so stable that they are unrideable. There was public TV show on bicycling about 10-15 years ago
which had film of Wilson's riderless bike sailing down a long gradual descent.

We ride bikes- in part- by making minute corrections, steering the bike under us as we fall to the
side. When we are first learning to ride, those corrections are large and we weave back and forth;
once we are proficient, the corrections are mostly unobservable even to ourselves and we appear to
ride in a straight line.

 

[Jasper Janssen]

On Tue, 17 Jun 2003 21:26:28 +0100, "Tim Cain" <tim_no1@you_know_what_to_cut_timcain.co.uk> wrote:

>Bikes are stable and self-correcting already.

Uh, no. A bicycle is an inherently unstable system requiring constant adjustment by the rider. This
is why robot bicycles (See the yearly BBC event Techno Games) are a fairly big challenge, and why
children can't intuitively ride a bike, like they can intuitively be a passenger in a car.

Jasper

 

[Jasper Janssen]

On Tue, 17 Jun 2003 15:53:05 GMT, "Jacobe Hazzard" <[email protected]> wrote:

>I furthered the idea somewhat, what if the gyroscope had an excess of momentum, then the net
>rotational inertia would be out to the left and the bike would lean *out* of turns!

No it wouldn't. Bike leaning has nothing to with gyroscopy. As a matter of very simple high school
freshman physcis, the line from the center of gravity of any object that isn't falling over in the
direction of the net forces acting on that object must go through the polygon formed by the contact
points with the ground. In the case of a two wheeler, the COG must lie above the line between the
two tire contact patches. When you make a turn, via the same basic physics, you are experiencing an
acceleration directed inwards of your turning circle. You can only have that acceleration when the
net force gets a non-zero inward component, and that means in turn that you need to shift the center
of gravity inward.

The gyroscopic effect is *purely* and only the effect that an object with rotational momentum will
have fairly significant resistance to changing the plane of said rotation. The effect is too small
to keep your bike upright, and it certainly doesn't affect how far you have to lean in corners.

It's very easy to demonstrate that steering is what keeps a bike upright: just ride no-hands for a
bit. You'll note that immediately it becomes harder to go in a straight line (nowhere near
impossible, just harder), and you'll also note that you can steer by shifting your weight around,
which you need to do more of to stay upright, and thus you will notice more.

Jasper

 

[Ron Hardin]

Jasper Janssen wrote:
> The gyroscopic effect is *purely* and only the effect that an object with rotational momentum will
> have fairly significant resistance to changing the plane of said rotation. The effect is too small
> to keep your bike upright, and it certainly doesn't affect how far you have to lean in corners.

Not only is it small, it's zero. Park your bike with the rear wheel propped up by the chain stays,
and spin the rear wheel as fast as you can. Now push the bike over. It falls as fast as it does with
the wheel stationary.
--
Ron Hardin [email protected]

On the internet, nobody knows you're a jerk.

 

[Squid-In-Traini]

> Litespeed already sells such a bike. High tech super lightweight full titanium construction, Price
> is only $850
>
> http://216.247.25.241/miva/merchant.mv? Screen=PROD&Store_Code=LS&Product_Code=T

I think you bring this up anytime a tricycle is mentioned. Did you buy too many of these and want to
get rid of some?

--
Phil, Squid-in-Training

 

[Tim McNamara]

In article <[email protected]>, Jasper Janssen <[email protected]> wrote:

> On Tue, 17 Jun 2003 21:26:28 +0100, "Tim Cain" <tim_no1@you_know_what_to_cut_timcain.co.uk> wrote:
>
> >Bikes are stable and self-correcting already.
>
> Uh, no. A bicycle is an inherently unstable system requiring constant adjustment by the rider.
> This is why robot bicycles (See the yearly BBC event Techno Games) are a fairly big challenge, and
> why children can't intuitively ride a bike, like they can intuitively be a passenger in a car.

Children can't intuitively walk, either. While balancing on two contact points is natural to humans,
the motor skills and perceptual sensitivity have to be learned not intuited. Being passively
supported by the environment- as in a car- takes no intuition or learning at all. A half gallon of
milk can do it too, as I proved on my way home from the co-op.

Bikes can be designed to be self-correcting and stable without a rider, but for all intents and
purposes they are so stable that they are unrideable. You can't get them to turn. They can be so
unstable as to be unrideable. Good bikes fall in between these extremes- unfortunately most people
are told by bike media pundits that they should be on the most unstable bike they can keep upright
because it's "more responsive."

As an aside, there's an interesting bike design in which the cyclist sits sideways between the
wheels, balancing front-to-back rather than left-to-right, mentioned in the most recent VeloVision.
I think it would be quite an adjustment to learn to ride.

 

[Squid-In-Traini]

"Tim McNamara" <[email protected]> wrote in message
news:[email protected]...
> In article <[email protected]>, "Tim Cain"
> <tim_no1@you_know_what_to_cut_timcain.co.uk> wrote:
>
> > Bikes are stable and self-correcting already.
>
> If that was the case, your bike could ride without you. It would coast, riderless, without
> falling over.

And that's what happens. Check this video out:

http://plaza.ufl.edu/phillee/****/ghostridingbike.avi

--
Phil, Squid-in-Training

 

[Jasper Janssen]

On Tue, 17 Jun 2003 22:13:12 -0500, Tim McNamara <[email protected]> wrote:

>Bikes can be designed to be self-correcting and stable without a rider, but for all intents and
>purposes they are so stable that they are unrideable. You can't get them to turn. They can be so
>unstable as to be unrideable. Good bikes fall in between these extremes-

Either of those pretty much are not bicycles, though. They're special devices that bear a passing
resemblance to bicycles except that the steering arrangement is wrong.

>unfortunately most people are told by bike media pundits that they should be on the most unstable
>bike they can keep upright because it's "more responsive."

Well, if you really feel you need to be able to make a 90 degree turn with a radius less than a
meter. Personally, I feel fine with relaxed geometry, and have no intentions of (deliberately)
riding anything more unstable. Of course, if I want a particular type of bike, I'll be forced to it,
pretty much, aside from custom frames or looking very hard for the one frame that fits my need.
Probably will do the latter next time I have cash to spare.

Jasper

 

[Jasper Janssen]

On Tue, 17 Jun 2003 22:14:50 -0400, Ron Hardin <[email protected]> wrote:

>Not only is it small, it's zero. Park your bike with the rear wheel propped up by the chain stays,
>and spin the rear wheel as fast as you can. Now push the bike over. It falls as fast as it does
>with the wheel stationary.

It's not zero, it's very, very small. Not detectable with inaccurate measurement instruments like
the unaided human brain. Take a wheel, out of the bike, hold by the axle, spin, and then make the
translation from spinning vertically to horizontally. You will notice that this is much more
difficult than when the wheel is not spinning.

The magnitude of this stability remains the same whether or not the wheel is in the bike or not --
it's just that the forces involved are completely disparate. Certainly way under 10% difference for
your experiment, and you really won't detect anything like that by looking at it. I suspect that the
effect might be smaller than the difference in starting positions, by the way, making it impossible
to detect unless you make a rig.

Jasper

 

[Ron Hardin]

Jasper Janssen wrote:
> >Not only is it small, it's zero. Park your bike with the rear wheel propped up by the chain
> >stays, and spin the rear wheel as fast as you can. Now push the bike over. It falls as fast as it
> >does with the wheel stationary.
>
> It's not zero, it's very, very small. Not detectable with inaccurate measurement instruments like
> the unaided human brain. Take a wheel, out of the bike, hold by the axle, spin, and then make the
> translation from spinning vertically to horizontally. You will notice that this is much more
> difficult than when the wheel is not spinning.
>
> The magnitude of this stability remains the same whether or not the wheel is in the bike or not --
> it's just that the forces involved are completely disparate. Certainly way under 10% difference
> for your experiment, and you really won't detect anything like that by looking at it. I suspect
> that the effect might be smaller than the difference in starting positions, by the way, making it
> impossible to detect unless you make a rig.

There is not the slightest difficulty in minutely and accurately changing the angle of a helicopter
rotor blade, in spite of its being huge and fast. You may be interested to hear that to tilt it
forwards, you apply lift 90 degrees away from the point you want to lift. Do that with your bicycle
wheel and you'll notice it moves easily.

Alternatively, _prevent_ it from turning 90 degrees away (say by using the rear wheel in the frame)
and the bearings will provide exactly the force necessary to move the wheel exactly as you pushed
it. The 90 degree effect disappears completely.
--
Ron Hardin [email protected]

On the internet, nobody knows you're a jerk.

 

[Tim Cain]

"Tim McNamara" <[email protected]> wrote in message
news:[email protected]...
> In article <[email protected]>, "Tim Cain"
> <tim_no1@you_know_what_to_cut_timcain.co.uk> wrote:
>
> > Bikes are stable and self-correcting already.
>
> If that was the case, your bike could ride without you. It would coast, riderless, without
> falling over.
>

They do.

Try it sometime - a sloping parking lot is your best bet. Point the bike downhill, give it a good
shove, and off she goes. (Obviously, don't use your Sunday-best Bianchi for this kind of stunt).

Best,

Tim.

---
Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.490 / Virus Database: 289 - Release Date: 16/06/03