Python code written for the new Drake distro.
Before using, source setup.sh. Then python setup-test.py to make sure pydrake and mathematical programming work correctly.
- Create basic 2D footstep planner that can create a path from a goal position to end position in unconstrained space
- Make the reachable regions from footsteps into general convex regions
- Add a nominal and reachable distinction (so it is much costlier to step outside of some scaled-down version of the reachable region)
- Add an option to make the reachable region into multiple convex regions
- Use LCM to feed plans into the old MATLAB simulator (
runAtlasWalking.m) to allow for simulation of created footstep plans.
- Tune constants to make footstep plans reasonable for Atlas
- Link in COUENNE Solver to solve angle constraints
A Python package that can create footstep plans based on various input parameters, including start position, end position, convex obstacle-free regions, and nominal footstep regions. See the README in the folder for more info.
Can solve for the closest point within a number of 2D regions to a goal point (run with -s to save the outputted image).
Can solve for the closest point within a number of 3D regions to a goal point (run with -s to save the outputted image).
