Star Wars VII Soccer Ball Droid

[Edwardowan]

A smooth ball on sand can't propel itself forward because of lack of traction. Think tire without profile.
The droid IMO needs to have some decent weight to be stable and reach the speed that is shown. Otherwise it would bump around and jump up into the air on uneven ground.

its entirely possible the practical effect BB-8 RC model was shot on an indoor/outdoor stage with blue or green screen then digitally composited onto the tatooine location scene much like the snow speeders were added to the Hoth flyover, the speeder bikes to the redwood forests and X-wings over the death star, etc. adding speed and terrain bump and bounce to the rolling action would be easy in post and Filming on a closed Pinewood lot stage would solve the problem of leaked photos of the thing as well as protecting it from the elements.

 

[DaddyfromNaboo]

its entirely possible the practical effect BB-8 RC model was shot on an indoor/outdoor stage with blue or green screen then digitally composited onto the tatooine location scene much like the snow speeders were added to the Hoth flyover, the speeder bikes to the redwood forests and X-wings over the death star, etc. ...

Yoz are decribing classical model work and vfx conpositing.
R2 was and is a practical effect. Jabba was a puppet. The door droid at Jabbas palace was a practixal effect. Treadwell droid and the mouse droid were practical effects. No postwork except from the sound dept.

Back on topic. I find xrobots find the most convincing. to date. What i can't understand is the actual material of the ball. A hard shell would make the ball bump around even more at high speeds and rotations. A softer material that even provides grip would be ...?

I can see a successor to the magic yoda 8 ball....

 

[Inquisitor Peregrinus]

I have been thinking (and reading) about this and have to agree this is either a split design of some sort or a combo practical/CGI.

If it uses magnets for the floating head, I don't think it could stay on with the forward momentum of the body. Floating by magnetism uses an equal push/pull force, so there is not to keep it in the same location as the body, which is being propelled forward.

Assuming magnetic levitation rather than bearings, if the body is just spinning between the poles of the magnets (one in the base of the head and one just inside the body shell at the top of whatever internal support), it'd look a lot like what we see. Rare-earth or electromagnets provide plenty of attractive force, both direct and shearing.

Also, shouldn't there be some sand being kicked up by the droid at the speed it's going? It doesn't appear to be a completely smooth rolling surface (the droid), which could keep the dust up to a minimum.

I's not a direct-drive wheel. It's basically the inner mechanism -- which does have direct-drive wheels with traction -- rolling up the inner surface of the ball and overbalancing it in that direction. The droid is basically just constantly "falling" in the direction of motion. No traction required, and thus nothing being kicked up behind it.

Edit: if all other problems were solved and it had a floating head, turning the head could be done with a small fan on each side or compressed air. These could be used like thrusters.

Or, more likely, by having the body frame end of the mag-lev array mounted on a swivel base. The frame will be heavy enough to resist Newton's Third Law, where the head probably isn't.

Regarding the Omni Wheels....? The front and back propel the unit fore and aft. But wouldnt the left and right 'Omni Wheels' create drag on the hamster ball? I will admit, this concept does address the floating head mystery.

I haven't looked far enough into it yet, but I know with bots like this that has been addressed. Got a local award-winning high school robotics team I could ask. *heh*

And I gotta give props to a guy who has been on here longer than I have and only posted 252 posts. There is something to say about he who speaks the least, speaks the loudest.

I'm one of those weirdos who has never really worried about post count. I just only try to say something when I have something to say. Which... I admit sometimes results in mini-essays rather than short answers, but I at least try to be interesting, so... *shrug*

A smooth ball on sand can't propel itself forward because of lack of traction. Think tire without profile.
The droid IMO needs to have some decent weight to be stable and reach the speed that is shown. Otherwise it would bump around and jump up into the air on uneven ground.

A metal internal frame, battery (especially battery), and control unit will definitely add some heft, and below the centerline will help stabilize. As for the traction thing, see earlier in his post -- the ball body is not a direct-drive wheel. The inner mechanism is climbing the inside of the ball body to push it forward as it overbalances in that direction. No traction required on the outside.

--Jonah

 

[Johnnymuffintop]

I just realized, JJ is exceptionally good at blending practical and CG. His work on Star Trek had a lot of aliens designed (not all were in the final movie of course) that were mostly practical with a digital twist. It's quite possible that to get the shot of BB-8 they dangled the head on a line like a puppet and digitally added in the ball. Won't know for sure until more clips, if any, come out.

 

[MrDynamite]

The BB-8 in the trailer shots are not a practical effect. full stop.

 

[SupaTrooper]

This CGI artist Julien Leterrier should get a job with ILM or Pixar for this amazing 3D render

Droid modeling / rigging / texturing / animation / lighting & rendering / compositing & film emulation
4 days to do it

[video=vimeo;113795338]http://vimeo.com/113795338[/video]

 

[AndyJ]

Here's another idea inspired by the CG Vs practical debate. This one would only work on-screen with CG enhancements...



Basically it's a solid "tail" that freely spins on the axle, with a motorized drive wheel on the back. The tail itself allows mounting of the RC electronics and servo to steer the wheel. The post the head sits on is combined with the tail so it's always upright. The hemispheres are allowed to spin freely, so they just rotate from the friction against the floor (or perhaps their own motor as well, just for show perhaps.)

Again...obviously not going to look good in real life. But with simple CG removal of the tail, it could look good on film.

 

[IAmTheClayman]

Not sure if anyone's linked to this article yet, but Mark Hamill confirmed that BB-8 (apparently the little guy is a BasketBall and not a SoccerBall droid) is a practical effect. And apparently hardy enough that they let him drive it around a bit

http://www.ign.com/articles/2014/12...wars-the-force-awakens-bb-8-is-a-prop-not-cgi

 

[Inquisitor Peregrinus]

Yep. People are still arguing the point, though.

--Jonah

 

[DaddyfromNaboo]

Not sure if anyone's linked to this article yet...

Rule of thumb: if it´s news older than a day, then it very likely has been posted and a search before posting may be in order Just one page back you´ll find a link about this.

Regarding the ball being driven by an inner source and not a direct drive motor, I am sure that a smooth ball without enough tracking will slip away under the "r/c car" if it goes high enough to topple it over. Think hamster in a ball on smooth surface

I know there´s an edge keeping the ball from falling off the plate, but the porcelain is so smooth that I bet he wouldn´t have made it far anyway.

Small spots of rubber may help give traction.

 

[Krats]

We used to have a tiled floor in our kitchen, and the school hamster was more than capable of propelling his ball around at great speed. I don’t think friction would be an issue unless great acceleration was involved.

Since this thread was raised I have been considering how this effect might be achieved. The information from Mark Hamill confirms this is achievable as practical effect, operable by a lone non-specialist.

Obviously the Segway based solution would involve far too narrow a point of contact to allow for stability. Whilst it is still possible this might be a practical effect with CGI enhancement I believe the design Inquisitor Peregrinus posted back on page 4 is both elegant and feasible.

 

[NeoRutty]

Just listening to Kevin Smith - Hollywood Babbleon... he was saying when he was on set BB-8 was there and the "controller" was a puppeteer... done with bars and wires.

"I talked to the puppeteer... who controls BB-8. Now this I was like 'really? puppeteer?'. Like you look in that trailer and you're like 'ohhh here comes a ****ing CG...'... It's a puppet. It's a dude with bars and ****. And he's like 'generally it's not even 'we've gotta green screen the bars out'... it's like we're just trying to make it tactile'"

http://smodcast.com/episodes/december-12-2014/

1:30 or so in...

 

[mvanwalks421]

Maybe for some scenes, I imagine; but I think it's most likely just one of many ways to control the BB-8 droid.

 

[Formori]

Yep. People are still arguing the point, though. --Jonah

Jonah, I tend to agree with you that it sounds like (from all reports) that it is a real RC robot and not just a mashup of practical and CGI but looking at many of the scenes from the prequels and the new Star Trek movies they used a lot of practical prop pieces while adding details or functionality with CGI.

When I first saw BB-8 on the trailer I couldn't not think of Sphero (I have one) and even though the body is smooth acrylic he still picks up tons of traction on wood and tile floors because the inner mechanism forces enough weight forward of itself and will gain momentum very quickly but sharp turns slide out a lot and it's not precise. That being said, a much bigger (and much heavier) robot would have significantly more weight to throw around inside the ball and with the right control system (omni-wheels inside the ball that are gyro stabilized like a quadcopter) he should have no problem quickly gaining speed on any surface and changing direction.

As for the head floating on top, how about a simpler system where a mast inside the ball pushes against the top-inside of the ball with caster bearings, holding strong magnets just under the rolling surface and the same bearings under the dome with magnets just above the ball surface? The magnets would be more than strong enough to keep it in place and not bouncing off while the roller-bearings keep the friction minimal and his head would only tilt as much as the interior mechanism (which on my sphero isn't much and should be less on a larger ball). The weight of the battery, control system, and frame would be more than enough weight to keep the ball stable and efficient.

For a practical droid (like all of the R2's that the Astromech guys build) this would more than suffuice even if we can't make tools come out and whatnot.

Another quick thought, that using reverse-polarized magnets in an on/off pattern would hold the head in an orientation, and if you were to swivel the mast under the ball surface the head should follow giving unlimited horizontal pan movement while keeping the head mounted to the ball with a magnetic connection instead of a physical one.

For those of you still doubting that a ball droid has any practicality, I can get my sphero to jump ramps and then down stairs in my house without damage, and his manoeuvrability is amazing for a little RC system. It would be much more efficient than R2's legs or any wheeled system (with the exception of a unicycle-styled robot, which is also proven tech and actualy has more mobility than the sphero but at a much higher cost to implement and when he shuts down he falls over).

I'm curious enough now that I want to buy a hamster ball and try out some omni-wheels inside with a basic stabilized control system (I built a mini-segway a few years ago, the stabilization is nearly identical). Let me know what you guys think!

 

[DaddyfromNaboo]

Jonah, I tend to agree with you that it sounds like (from all reports) that it is a real RC robot and not just a mashup of practical and CGI but looking at many of the scenes from the prequels and the new Star Trek movies they used a lot of practical prop pieces while adding details or functionality with CGI.

When I first saw BB-8 on the trailer I couldn't not think of Sphero (I have one) and even though the body is smooth acrylic he still picks up tons of traction on wood and tile floors because the inner mechanism forces enough weight forward of itself and will gain momentum very quickly but sharp turns slide out a lot and it's not precise. That being said, a much bigger (and much heavier) robot would have significantly more weight to throw around inside the ball and with the right control system (omni-wheels inside the ball that are gyro stabilized like a quadcopter) he should have no problem quickly gaining speed on any surface and changing direction.
....

I'd say more mass means more inertia to counter. I am not sure if that would help woth maneuverability.

I'm curious enough now that I want to buy a hamster ball and try out some omni-wheels inside with a basic stabilized control system (I built a mini-segway a few years ago, the stabilization is nearly identical). Let me know what you guys think!

You have to ask?!

 

[Formori]

I'd say more mass means more inertia to counter. I am not sure if that would help woth maneuverability.

The problem that my sphero has when trying to turn at high speed is that the ball surface will turn in the correct direction but because of his momentum he keeps going forward and the ball slips, so he kind-of drifts around. At low speed the ball has enough forward weight to overcome the momentum and shift him sideways without slipping, so he can go from one direction to another directly without actually turning, which is why I think the bigger ball will make a difference (ie. the gyroscopic effect a motorcycle wheel has compared to a bicycle wheel). The BB-8 droid will never travel faster (mine anyway) than my sphero which rockets around and bangs into stuff and so the larger ball and weight transfer should equal more control and the ability to corner better. This could also be improved with better gyro stabilization in the motor controller to adjust for the amount of turn versus the amount of tilt the mechanism has to do in order to move the ball.

You have to ask?!

Well then, I need to go get me a hamster ball and dig out the old robot to tear him apart and see what I can do.

A question to those here familiar with movie shooting and the existing droids in the series: How big do you think BB really is?

After screen-capping the trailer and trying to scale from his surroundings I think he's roughly 18" tall which would be a 12" diameter sphere and a 6" diameter head, which seems to be the rough agreement on the Astromech forums also. I may start with a smaller version to make it easier to work with and then scale him up to a proper size later (when we have more info) but I'm trying to wrap my head around the dynamics of a large sphere moving around in an enclosed space (like a convention hall) and how fast I should make him move (not how fast he could be because it's a giant wheel, it could be ridiculously fast and uncontrollable).

Thoughts?

 

[Inquisitor Peregrinus]

Okay, the idea of a drifting droid kinda makes me happy. *chuckle*

As for the head floating on top, how about a simpler system where a mast inside the ball pushes against the top-inside of the ball with caster bearings, holding strong magnets just under the rolling surface and the same bearings under the dome with magnets just above the ball surface? The magnets would be more than strong enough to keep it in place and not bouncing off while the roller-bearings keep the friction minimal and his head would only tilt as much as the interior mechanism (which on my sphero isn't much and should be less on a larger ball). The weight of the battery, control system, and frame would be more than enough weight to keep the ball stable and efficient.

Another quick thought, that using reverse-polarized magnets in an on/off pattern would hold the head in an orientation, and if you were to swivel the mast under the ball surface the head should follow giving unlimited horizontal pan movement while keeping the head mounted to the ball with a magnetic connection instead of a physical one.

That's what's shown in the XKCD blueprint I posted, and what I proposed for turning the head.

For those of you still doubting that a ball droid has any practicality, I can get my sphero to jump ramps and then down stairs in my house without damage, and his manoeuvrability is amazing for a little RC system. It would be much more efficient than R2's legs or any wheeled system (with the exception of a unicycle-styled robot, which is also proven tech and actualy has more mobility than the sphero but at a much higher cost to implement and when he shuts down he falls over).

I imagine the only thing R2 could do better is climb stairs.

--Jonah

 

[Formori]

I also just realized that with the sphero version of propulsion whenever he goes to take off quickly his head will be tilted back making it look like he's barely hanging on, should give him quite a bit of personality.

I'm going to have a dig through my parts boxes tonight to try and set up a prototype to see how it might work.

 

[kczimmy]

 

[joez2391]

what?!?! I can't believe someone built a BB-8 already, great job it looks awesome!