Notes from the PhD coursework at SUTD: what each course covered in general terms, what I built, and how it connects to the thesis threads (robust robotics, multi-robot perception and long-term autonomy). Course material and assignments stay private. The courses page has the full list, including the ones from long ago.
- Multimodal Generative AI: transformers, contrastive alignment, diffusion and grounding. The project became GoGet.
- Systems Security: memory safety, web trust, DNS, side channels, and a team project on attacking a robot’s middleware.
- Secure Cyber-Physical Systems: control system security on a water treatment testbed, from both the attacker’s and the defender’s side.
- Foundations of Robotics and Multi-Robot Systems: control, planning, estimation and coordination, and a search and rescue run with a team of robots.
- Reinforcement Learning for Embodied AI: from tabular methods to self-play, and Paddle Duel.
- Networked Embodied AI: vision-language-action models and the network underneath them.
- Robotics Enterprise: taking a robot from prototype to product.