Procedural Locomotion?

I was looking over the PhAT the other day, and it occurred to me that it might be possible to create a procedural locomotion system, based on a number of options and parameters I saw in there, and a number of blueprint nodes I’ve run across in my misadventures with the blueprint system.

Now I’m curious: Has anyone ever tried to implement a procedural locomotion system through blueprints before? Is it even possible to do so, or do I just have the wrong idea about it could be done that way (and it would have to be done mostly, if not entirely, through C++?) It’d be really great if it could be done through blueprints (since I’m complete beginner in C++) but I’m not holding my breath for it.

You can’t control PHAT constraints with blueprints. And for some reason PHAT doesn’t let you see the actual transform data for things, so it’s a real pain to do anything accurately. But you can do the same things outside of PHAT with regular old physics constraint components (or actors). You don’t have the convenient interface for testing things, but it’s the same stuff. I did read a thread where some guy talked about building self-balancing and controllable “robots” with physics constraints, but I guess he removed his profile or something, because all of his posts are gone. :confused: So, if I’m understanding correctly what you mean by “procedural locomotion”, then yes, it’s possible, and with blueprints. The hard part is knowing how to calculate what forces to apply, etc. (and dang euler angles)

Yeah it would be possible in blueprints by using sockets and physics attachments to pull on individual parts of the body but you wouldn’t need to be using PHAT for that.