Welcome to the Robotics Course Repo¶
Github repo: https://github.com/MarcToussaint/robotics-course
This repository is used in three lectures:
- The Introduction To Robotics lecture at U Stuttgart
- The Lab Course with the Real Baxter we offered for some years at U Stuttgart
- And now the Lab Course using simulation at both, TU Berlin and U Stuttgart
They all share the same core code, but have quite different installation setups and exercises. Embedded below are exercises for the lab course in simulation, and generic documentation of the code. The code documentation refers to python interfaces – but these are almost identical to how users should use C++.
- Lab Course in Simulation
- Robotics Code Overview
- Learning KOMO (K-Order Markov Optimization)
- No dynamics? No velocities? How do you solve for dynamic problems?
- What is KOMO?
- What are the main steps to define a KOMO problem?
- Roughly, what does KOMO do with these specs?
- Can KOMO only optimize paths?
- Can I load an URDF?
- First practical steps
- Features are at the core
- Varying decision variables per time slice: kinematic switches & force contacts
- Summary: there are only a few fundamental methods in KOMO, the rest are helpers
- Steps, phase, real time, and time optimization
- LINKS to more documentation