Intelligent Systems for Robotics · ten teaching weeks
The course runs bottom-up: skills first, then the knowledge representation that describes them, then planning and execution over both. Weeks 1 to 8 are lectures, each with a deck, notes, a runnable example, and a micro-project. Week 9 is project work and week 10 is presentations.
| Week | Topic | Deck | Notes | Micro-project |
|---|---|---|---|---|
| 1 | Course introduction; the layered architecture; the fetch scenario | Deck | Notes | exercises/01-introduction/ |
| 2 | The skill model: options, initiation sets, termination. RL fundamentals | — | — | — |
| 3 | Training robot policies in MuJoCo; transfer to Webots | — | — | — |
| 4 | Ontologies, RDF/OWL, description logics; modelling the apartment | — | — | — |
| 5 | The topological map and meta-actions in the KG; grounding detections | — | — | — |
| 6 | Classical planning: STRIPS, PDDL, state-space search, heuristics | — | — | — |
| 7 | Planning over skills and meta-actions; hierarchy and plan quality | — | — | — |
| 8 | Execution and monitoring; behaviour trees; blocked edge to replan | — | — | — |
| 9 | Project work — no lecture | — | — | — |
| 10 | Project presentations | — | — | — |
Each micro-project runs from the inputs it provides, so a week missed does not block the next one. Connecting the pieces into one working system is the final project, and that integration is what is assessed.
Intelligent Systems for Robotics · course materials