Hacker News Reader

Comment by gcb0 | Hacker News Reader

and no word on how it moves. even bbc is going downhill :(

It seems to be rotating via reaction wheels (via this[0]). I haven't yet figured out how it translates, though.

EDIT: if you have gravity available, the video on that site illustrates a clever way to use those reaction wheels to move around by tumbling.

EDIT2: "for maneuvering around in space there are twelve electric micro-fans or μFans"[1]. The SE thread has a screenshot of a video which shows the placement of those fans.

EDIT3: Full video about the Int-Ball[2].

[0] - http://www.kenkai.jaxa.jp/eng/research/electrical/triaxial.h...

[1] - https://space.stackexchange.com/questions/22324/how-does-jax...

[2] - https://www.youtube.com/watch?v=ZtIARUS7Lqc&feature=youtu.be...

From a related video: the reaction wheels spin up while keeping total angular momentum at zero, then apply a rapid (magnetic) brake to transfer the momentum to the body (from the wheels). I'm assuming if you do that precisely enough (and in three dimensions), you can generate linear movement.

EDIT: I'm probably very wrong about the movement.

EDIT: The video: https://youtu.be/n_6p-1J551Y

how does conservation of linear momentum apply? edit: this question was asked when the parent comment claimed that this machine could "generate linear momentum" from the wheels.

Imagine holding up a bicycle in the air from the midpoint. Now spin both wheels in opposite directions. No movement. Now apply one brake. Now you have movement. Now do that in three dimensions with precision and you have this thing.

Conservation of momentum applies when there are "no external forces", which wouldn't be the case here when you apply the brake.

EDIT: Yeah, my apologies. I didn't mean to write "momentum".

The brake is an internal force though (the force and its reaction both act on the bicycle). The whole unit's only external forces are from the fans so that's all it can use to translate. That and bumping into things.

> That and bumping into things.

Come to think of it - in air you're constantly bumping into things - air molecules. I think it could technically be possible to steer in zero-g within an atmosphere by pure rotation (e.g. consider that a fast-spinning object is kind of a (crappy) fan).

And, in fact, that's what this device does.

It uses an array of 12 small rotating fans placed around its surface to push jets of air and move itself from place to place.

Yes, I know.

My point was that in principle, you could abuse reaction wheels to give you translation capability if you're moving through air.

Can it face you with one of its fans and run it at the same high speed as the opposite fan, to keep you cool, and take photos of you with your hair blowing in the wind?

though less technical, "movement" implies "momentum", no?

In the person + bike system, the brake action would result in internal forces. I don't see how this system could translate.

Not sure about the movement-momentum question.

Thinking about it, the bicycle would only rotate about the center unless it had something to react against, as was pointed out by others.

well, momentum is defined as the product of mass and velocity. movement implies velocity, and we are not talking about massless things here, so yes, movement implies momentum.

The system in the (cool) video is in contact with a table.

The table exerts a "normal" force on the cube, allowing a change in vertical momentum. If the system in question is the cube alone, this is an external force.

This method could be used if the drone were cubic and near a wall of the space-station, by kicking off of the wall. That could get it moving, but until it hit the opposite wall, there's nothing it could do to stop.

You're right. I was thinking about that, but I really don't know. This is pretty far out of my expertise. averagewall made a good point [1].

[1] https://news.ycombinator.com/item?id=14796338

>I'm assuming if you do that precisely enough (and in three dimensions), you can generate linear movement.

You can rotate it. You wouldn't be able push it around that way though.. if you manged that it would be a reactionless drive..

What if the wheels aren't centered? You could probably not get linear movement - but couldn't you rotate around a point other than the center?

..a...b..

Rotate ~180 degrees around a, then b?

[ed: from other comments I see I may be too used to thinking in a gravity constrained environment..?]

I'm not positive, but I'm pretty sure wherever you put the wheels, you aren't going to get the drone to rotate about any point other than its center of mass.

Makes sense, when thinking of it like a force - any where you push at the exterior (or interior) that's not directly at the center, would cause rotation around the center. I gather internal forces would add up/cancel out and never end up in a vector that pushes on the center of mass.

The use of fans is disappointing; that means it can't go outside. My first thought when seeing it is that it'd be great for doing external inspections, and could even have its own private little airlock to get in and out.

Maybe they could make a version with ionic thrusters, that stays outside all of the time. It could have a 'nest' where it parks and refuels when not in use.

I'd imagine that'll come in time.

Easier to start inside, where you can pick it up if it runs out of batteries or glitches out. Less chance of poking a hole in something critical, as well.

Ion thrusters have tiny acceleration, which is why they're best on deep space missions (slow and steady wins the race). I wouldn't think they'd be practical for this type of thing, which you'd want to go one or a few kilometers per hour over distances of tens of meters.

You could just set rails or clotheslines outside the station for an external drone camera to travel along.

Or maybe not whole rails, but just fixtures the drone could grab onto. You could supply the drone with power and control signals through the fixtures, and the drone could kinda travel everywhere by just grabbing onto a nearby fixture, then releasing the previous one. Wouldn't that be great? /s

Canadarm2, the robotic arm on the International Space Station, does this. https://en.wikipedia.org/wiki/Mobile_Servicing_System#Canada...

Let's add more Canadarms, and have them throw and catch the drone around.

I'm sure there are very strict rules about having anything flammable or explosive on board ISS, which pretty much means you can't have a drone that can move under its on propulsion both inside and outside the station.

Just power it with compressed air.

Expending air is not sustainable.

Maybe make it magnetic and roll it on the exterior hull.

Won't work in space.

Magnetism doesn't work in space?

away from the hull

NASA's Manned Maneuvering Units used compressed nitrogen. But I'm not sure there'd be a huge benefit to having indoor/outdoor capabilities in a single drone. You'd have to cycle the airlock every time it went in or out, and it would be bulkier and less efficient in either mode than a dedicated unit.

I wouldn't want to be on the wrong end of a small particle accelerator and I wouldn't want any part of my spaceship in that position either.

Here's a video that shows the actual thing moving in zero gravity:

https://www.youtube.com/watch?v=PntNBeJuRFQ

I also found this quote amusing: "The drone can float in a zero-gravity environment"

Here is an article straight from Jaxa: http://iss.jaxa.jp/en/kiboexp/news/170714_int_ball_en.html

They also don't say, but they link to this article: http://www.kenkai.jaxa.jp/eng/research/electrical/triaxial.h... which has a lot more technical detail. Unfortunately I don't have time to dig deeper, but hopefully it will help you search a bit better.

According to their press release it has 12 propellers (推進用ファン): http://issstream.tksc.jaxa.jp/iss2/press/170714_intball.pdf

Forgive the stupid question: what makes the use of reaction wheels and fans preferred over fans installed at different angles?

Energy efficiency or fan reliability perhaps?

Precision and speed, reaction wheels can go from start, rotate you one degree and then stop all motion. With fans your going to end up constantly correcting because you never hit zero rotation.

The robot is exposed to the station airflow so it has to adjust constantly anyways.

While non zero, the spherical shape and minimal ISS airflow generally makes this a non issue. You can use those reaction wheels to calibrate very precise adjustments over time, you can also just bump into something to get rid of excess rotation.

Note, space telescopes effectively use both approaches.

It may also minimize the impact all of that moving air has on everything else in the station. Without gravity, fans blowing your things around could get annoying.

Being able to operate with no atmosphere (e.g. outside of the ISS) would be an advantage I suppose.

Yet, the translation movements are done using fans. Perhaps it's a prototype and future version will operate outside.

Yup, I misread. I was thinking you meant what would be the benefit of not using fans at all.

You can get really good stability and precision using reaction wheels.

Any reason why it couldn't use air jet to create linear motion? It is floating in air after all.

Okay, just saw EDIT2 of TeMPOraL: https://news.ycombinator.com/item?id=14795857, it appears to move around using micro fans.

I was looking for the same thing. It has got a motor inside and uses some kind of momentum to rotate. Another motor makes it go forward, but at a snail's pace.

It has 3 motors for 3-axis reaction wheel set (for rotation), and 12 fans to move around.

Reply on news.ycombinator.com