hi all, I’m new to robotics and was wondering if anybody has thoughts on Position Based Dynamics methods: https://matthias-research.github.io/pages/publications/smallsteps.pdf. They seem to be able to account for constraints in a seemingly straightforward way.
if there is interest, I’m willing to prototype these algorithms in RigidBodyDynamics.jl, or start a new project altogether. im really just trying to get started with robotics simulations for kicks. my background is in writing fluid dynamics solvers. thanks in advance for your guidance
You might as well just use ModelingToolkit to tear into an ODAEProblem to handle the constraints and use a higher order adaptive solver which reuses Jacobians. That would be the same generalization to these techniques that Dymola uses, which would probably be a good order or two magnitudes faster than any 1st order technique (if you need more than minimal accuracy, if the events are not too too close, etc.).
I would really love to see Position Based Dynamics as a separate Julia library. I’ve been following the work of Müller, Macklin et al. for some time now and am enjoying Müller’s new Ten Minute Physics on Youtube. I am very new to Julia and am learning it specifically to speed up calculations for some future simulation projects. A new PBD Julia library would be incredibly useful not only for me but many, many others as well. The thing I am still trying to figure out for my simulations is how to best display the 3D output. Having Julia integrated into Blender, the Godot engine or even Raylib would be simply awesome. Anyway, I for one, fully support your idea to write a Position Based Dynamics library for Julia.
About Raylib and Julia… maybe you should have a look at Clang vs CBindingGen - #22 by mhinsch