Testing Kinematics with Papervision3D, Collada
A forward kinematics experiment using Papervision3D v2.0 alpha and a Collada 3D model of a robotic arm.
Hope you enjoy it.
I don’t want to release any source from this yet, as I want to do some more work on it, relating to the following items…
- As you’ll see, the range of rotation of each joint is constrained, but there isn’t enough program logic to keep consecutive joints from being rotated in such a way that the model folds in on itself unrealistically. Incidentally, the kind of model used here is ideal for playing with kinematics, as each joint rotates around only one axis, and each segment is connected on each end by no more than one other segment, except for the ‘pincers’. (Next up: bipeds . . . ?!)
- There’s some ‘Z-popup’ which could be mitigated by modifying the source model just a little. By the way, the model comes from Learning 3ds Max 2008 Foundation.
- The user interface to this piece is bare-bones functional. But I’m glad to have discovered that it’s possible to detect overlapping key-down events in Flash, which is kind of interesting. (If anyone has ideas for an easier-to-use interface that stills allows you to control multiple joints simultaneously, please let me know)
- Lastly, it makes sense at this point to (learn how to) do some basic inverse kinematics. If and when I figure something out, I’ll post an update…
January 30th, 2008 at 5:06 am
Love it !
The controls are working great, however i would replace 1q2w with the arrows key and for the rest like: pageup/pagedown home end last 4 maybe some sort of wasd setup. Or maybe we should pickup some special gloves for this :)
January 30th, 2008 at 9:39 am
Damn this would be fun to hook up to an Arduino. You could use potentiometers/digital compass/etc. to control the joints. Looking forward to your release of the source!
January 30th, 2008 at 10:10 am
Awesome! Great work!
January 30th, 2008 at 11:14 am
Shtoopid! No, just kidding. Looking great. I can’t wait to see how this can apply on some fancy sites.
January 30th, 2008 at 1:11 pm
i could play with it a whole day:D
January 30th, 2008 at 2:24 pm
Next step: Placing objects to pick up and interact with + physics engine that makes the objects bounce when they’re dropped again.
January 30th, 2008 at 2:57 pm
amazing!
i love it
January 30th, 2008 at 7:48 pm
fantastic work
February 1st, 2008 at 7:24 am
nice one.
cheers
_frank
February 3rd, 2008 at 10:11 am
Holy mackerel. That’s awesome.
I’d love to see some inverse kinematics going on that. I’m imagining a lot of flash stuff in the future with robotic arms and IK animations…
February 3rd, 2008 at 8:13 pm
Amazing!, Very Good!
one cool “next step”: aply bones on skin (Skinning) in papervision3d
http://gpwiki.org/index.php/OpenGL:Tutorials:Basic_Bones_System#How_does_a_bone_system_work.3F
February 5th, 2008 at 9:52 am
Absolutely incredible!
February 23rd, 2008 at 5:48 am
Great job … rly great
March 6th, 2008 at 1:50 am
It rocks !!!!
April 16th, 2008 at 4:32 pm
This is great, I just cant wait for IK
May 3rd, 2008 at 4:27 am
Jus rocking!! As a matter of thought, mouse movements could be associated for rotation!!
July 6th, 2008 at 12:45 am
This is really the art of new age media. I hope that this technology develops further so that it can be used regularly on sites.
July 25th, 2008 at 4:11 pm
Love it.
Someone needs to make a marionette with FK and IK support so it can be manipulated and also “jangle” and show the effects of gravity.
July 25th, 2008 at 4:40 pm
Totally. I’ve been thinking recently that even without IK-type intelligence, adding/implementing some properties/behaviors to each segment could yield some really interesting results. Angular velocity, mass, resistance, ‘bounciness’ … And gravity, as you said. As long as the effects were properly communicated up and down the chain of segments. That’d be fun…
Lee
August 10th, 2008 at 4:40 am
I really like this when are you likely to release the source.