Archive: 9 posts

---

I researched on how bipedal robots move in real life and was going to attempt to some how implement it into LBP. On a certain website it said I needed a good understand of a "balance gyro". Can somone tell me what this balance gyro is and is there any robot out there that has one that is similar?

2010-03-10 17:34:00 Author: Fumo161 Posts: 210

---

There are several ways to balance bipeds in LBP. It all really depends on what you want your contraption to accomplish. I had made a thread a long time ago detailing some methods I experimented with: https://lbpcentral.lbp-hub.com/index.php?t=8398-Bipedal-Motion-2-Balancing Also, a few members have made some gyro devices which you should check out: https://lbpcentral.lbp-hub.com/index.php?t=13030-SixAxiS-Engine%21 https://lbpcentral.lbp-hub.com/index.php?t=22150-Vehicle-tools-Tilt-and-motion-sensor-and-control-pod-%28Copyable%29 The ultimate point is that you want to keep your biped from tilting too far forward or too far backward (which would cause it to fall) by correcting its inclination somehow.

2010-03-10 18:02:00 Author: Gilgamesh Posts: 2536

---

My understanding of gyros is a bit fuzzy, but I believe it's a spinning wheel thing that tends to resist changes in inclination and can be used to detect those changes. It spins around a vertical axis, which is something that can't be done in lbp. So we had to come up with a different method of detecting inclination. The last thread that Gilgamesh linked to is the level I published showing a reasonably accurate tilt sensor (it does have some long-term accuracy problems, though, and it takes a bit of trial and error to get one working for the specific application you want). As far as I know, Dante95 never released his sixaxis sensor or detailed how it worked. In fact, that's pretty much why I made my tutorial level (I'm of the opinion that knowledge like that ought to be freely shared and since he wasn't sharing, I decided to figure it out the best I could and share it myself).

2010-03-10 21:25:00 Author: Sehven Posts: 2188

---

My understanding of gyros is a bit fuzzy, but I believe it's a spinning wheel thing that tends to resist changes in inclination and can be used to detect those changes. That's pretty much it. It's exactly the same technique which keeps a helicopter stable. It spins around a vertical axis, which is something that can't be done in lbp. ...which is another reason it's so awkward to build a helicopter in LBP. The last thread that Gilgamesh linked to is the level I published showing a reasonably accurate tilt sensor (it does have some long-term accuracy problems, though, and it takes a bit of trial and error to get one working for the specific application you want). I had a brief look at that, and I wasn't quite certain whether the frictionless bolt always prevented the pink floaty from changing its angle, relative to the ground, when exposed to sudden changes in velocity, or whether the embedded peach floaty was actually necessary or not.

2010-03-10 22:26:00 Author: Aya042 Posts: 2870

---

A while back I experimented with holding a block of material stable in the air. I ended up with an 8-point mag-switch detector ring and an emitter providing a dark-matter mounted mag-key. The detector ring fired one or more of eight rockets when necessary. It was a bit crude, and due to things being slightly off-centre and out of line I had to install a spin-suppression system - again using dark matter feeding signals to two spin correction retros. It was fun to watch, but completely impractical and useless - therefore ideal for LBP

2010-03-11 10:05:00 Author: Nakatsu Hime Posts: 59

---

I wasn't quite certain whether the frictionless bolt always prevented the pink floaty from changing its angle, relative to the ground, when exposed to sudden changes in velocity, or whether the embedded peach floaty was actually necessary or not. The tilt sensor can't handle sudden jarring motions, such as impact (meaning the vehicle's impact, not the sensor's--obviously an impact to the sensor would throw it off). I slapped one on a plane that TSFRJ built and it gave reasonably good results, but over time, even without any jarring impacts, it would drift off center and become useless. It was TheGide who discovered (and posted in that thread) that peach floaty could be used to correct the problem. It hadn't occurred to me before that, that peach floaty had approximately the same drag as pink. That said, I have had varying degrees of success with including peach floaty. The one I slapped on TSFRJ's plane worked perfectly, but I've had at least one that seemed to lean back (the peach floaty dragged more than the pink) so further experimentation is in order. I figure that if you were to make the entire top half peach, that would be bad, but if you don't use enough, it won't right itself, so perhaps a quarter of the overall area of the circle should be peach, or maybe a bit less even. I would also guess that having it closer to the middle of the circle (in the top half but concentrated toward the center rather than the edge) would probably work better since any dragging force it might exert would be closer to the axis than the rim. There may also be a particular size of disc that works best (due to the semi-loose physics of lbp) and a bigger or smaller one wouldn't work as well, but I haven't really tested for that either.

2010-03-11 20:26:00 Author: Sehven Posts: 2188

---

...but over time, even without any jarring impacts, it would drift off center and become useless. Ah. That's a shame. I tried playing with one for a bit, but couldn't get it to drift off-center. Was it possibly just colliding with another object in the thin layer? Anyways. I'll have to have another play with that. There may also be a particular size of disc that works best (due to the semi-loose physics of lbp) and a bigger or smaller one wouldn't work as well, but I haven't really tested for that either. Unfortunately, it's probably gonna be one of those things where no amount of tweaking will get it to work perfectly in all cases. Still, it's certain the best attempt I've seen so far.

2010-03-11 20:46:00 Author: Aya042 Posts: 2870

---

Yeah. Just wished I had thought of it. I got it from Asmadon: he was building a mech right around the same time I was and we got to comparing notes. He showed me the tilt sensor+dark matter emitter trick. You can see his demo if you do a search for his psn. I believe the level name was Raptor mech (there's two: the walking one uses the sensor and emitters, but the sprinting one uses rails to keep from falling over). The best thing about publishing that demo was how TheGide figured out the peach floaty trick: see how sharing things can end up coming back and helping you (I'm talking to all the people who like to keep their special tricks to themselves--Grrr). As for it losing its calibration, I never noticed the effect in my early experiments because I put it on a slow moving vehicle that didn't crash a lot. Stick it on something faster, especially something that makes frequent direction changes or hits a lot of bumps, and you'll probably notice that it starts to drift a bit. Also, the sensor HATES water (it loses its neutral buoyancy in water). Fumo, if you're still interested in knowing about gyros (even though it doesn't apply in lbp), here's a nice link I stumbled across: http://www.howstuffworks.com/gyroscope.htm Spinning wheels do have some other properties that I'm been meaning to experiment with in lbp but just haven't got around to. For example, when motorcycle racers jump, and want to bring the rear of the bike down while in mid-air, they'll rev the engine up and let the wheel spin (motorcycles are rear-wheel drive in case that's not obvious). I don't really understand why, but the spinning wheel causes the rear end of the bike to go down faster than the front, so they can land the jump without diving over the handlebars (since the front end leaves the ground before the rear end, it begins to fall sooner: which, combined with the heavier weight in the front end of most cars, is why making a car jump is generally a bad idea). Anyway, I've always meant to experiment with that: to see if spinning a wheel on a motorcycle in lbp will change the way it falls. Maybe I'll do that later tonight. I'm also curious if it'd be possible to "throw a curve ball" in lbp. Curve balls work because the ball is spinning: one side is moving faster than the other, therefore it tends to curve away from that side... or you could think of the balls spin causing it to "roll on the air." Since aerodynamics seem to be pretty much non-existant in lbp (drag in lbp, like when the wheels or legs of a tall vehicle try to roll out from under the top is caused by inertia, not wind resistance--that's another thing to take note of in mech building, Fumo) I don't know if spinning a wheel will cause it to arc when thrown in a straight line. Should be pretty easy to rig a launcher that can throw a spinning object in a straight line, so I'll have to try that too. Another interesting thing about spinning: I can't remember who it was or what kind of car they used, but there was a racing team way back when that managed to make their car significantly faster around a track even though modifications weren't allowed. The people in charge of the race took everything apart, trying to figure out what they had done, but couldn't find any modifications. Turned out, they had rigged the camshaft to spin the opposite direction from the stock setup, which allowed them to make tighter turns around the track (if the track's turns had gone the other direction: right instead of left; then the default cam direction would've been favorable). That probably doesn't have any lbp implications even if the other things do, but I always thought it was cool. [edit] Did my experiments. I don't think anything practical can be done with it, but the motorcycle wheel thing kinda' worked. Posted about it here. (https://lbpcentral.lbp-hub.com/index.php?t=23824-Aerodynamic-properties-of-wheels)

2010-03-12 04:44:00 Author: Sehven Posts: 2188

---

The best thing about publishing that demo was how TheGide figured out the peach floaty trick: see how sharing things can end up coming back and helping you (I'm talking to all the people who like to keep their special tricks to themselves--Grrr). I couldn't agree more. I'd never have come up with many of my ideas were it not for looking at other people's, like rtm's demitter demo, or the LBPC logic pack. Even oLMCo's music boxes inspired an adaptation to work around the limitations of synchronizing connectors. I've always thought it more sensible to publish your findings, rather than keeping them secret, since if it helps another creator to create an absolutely awesome level, which you subsequently play and enjoy, then surely that's a win for you. Worst case scenario is that they take the credit for your idea, but it's not as if you get payed royalties for such things, so what does it really matter? It's just vanity. Since aerodynamics seem to be pretty much non-existant in lbp... From my experiments, there's definately some sort of primitive simulation of atmospheric resistance going on. It's very apparent if you create a 1x1 thin block of polystyrene and a 1x1 thin block of metal, and let them fall to the ground at the same time. My guess is they just apply a constant force to all moving objects in the opposite direction to its velocity vector, and given that F=ma, then objects with a smaller mass will decelerate more quickly than ones with a larger mass. As for aerodynamics, I've seen no evidence that the shape of an object has any bearing on this calculation.

2010-03-12 12:56:00 Author: Aya042 Posts: 2870

Posts: 1077139    Threads: 69970    Members: 9661    Archive-Date: 2019-01-19

Aus dem Archiv wurden alle persönlichen Daten wie Name, Anschrift, Email etc. - aber auch sämtliche Inhalte wie z.B. persönliche Nachrichten - entfernt.
Die Nutzung dieser Webseite erfolgt ohne Speicherung personenbezogener Daten. Es werden keinerlei Cookies, Logs, 3rd-Party-Plugins etc. verwendet.