Working notes from my PhD. They change often and link to each other, and the graph shows the links. Finished pieces go to the letters. The diary is organised by robotics subfield, and every paper I read gets a note filed under one. How this diary works explains the structure.
Subfields
| Subfield | Covers |
|---|---|
| Perception | segmentation, depth, and the foundation models behind both |
| Mapping and localisation | SLAM, reconstruction, robust and certifiable estimation |
| Semantic mapping | 3D scene graphs, open-vocabulary maps, change over time |
| Planning and navigation | task planning, language-grounded planning, navigation |
| Control | barrier functions and safety filters |
| Robot learning | policies, vision-language-action models, benchmarks |
| Multi-robot systems | consensus, Byzantine teammates, collaborative SLAM. The main topic of the PhD |
| Safety and verification | runtime monitoring, formal methods, failure detection and recovery |
| Security | attacks on robots and their defences, cyber-physical security |
| Human-robot interaction | trust, interfaces, assistive robots |
| Field robotics | robots at sea, in the air and outdoors |
Thesis threads
Three directions my thesis might take, each cutting across several subfields: robust robotics, multi-robot multimodal perception and long-term autonomy. The thesis map puts them on one page you can pan and zoom.
Finding things
- Catalogue: every page, one line each, by subfield.
- Reading list: every paper as a table you can sort by year, subfield, domain and status, with the papers I haven’t read yet under To read.
- Entities: datasets, platforms, sensors, software, models and groups, each with the notes that mention them.
- Comparisons: open-vocabulary maps, failure detectors, defences against lying teammates.
- Log: what was added, and when.
- Course notes and dated entries.
- Explainers that go with the notes: Lying robots for resilient consensus, and 3D scene graphs.
Each note has a stage: seedling (a stub or an open question), budding (some structure and links, may still contain mistakes) or evergreen (reasonably complete). Hover over a link to preview it.
Notes are markdown, written in Obsidian and published with Quartz. Notes that aren’t ready are kept in a separate private repo.