For the Vienna Physics Engine, create a model for a bipedal walker, i.e. a human walking on 2 legs. You can use the rigid body and constraint implementation in the engine. Use motors to act on the limbs. Start with a 2D version and some starting estimation for the values, then use continuous optimization to create ML models for each motor depending on the state of the walker. You can use ANNs, reinforcement learning or any other ML method. You can also make use of any python library.
If you are interested in this topic or for more information please contact Prof. Helmut Hlavacs (helmut.hlavacs(at)univie.ac.at).