Notes, journals, travel logs, and reflections.

Final report: transferring the hybrid Isaac Lab and PPO system to a real Unitree G1 for bottle grasping and lifting.

Week 3: improving integrated success with object-relative control and moving the learned policy toward real G1 deployment.

Week 2: scaling to a 0.62 kg bottle while developing curriculum learning and a safer approach policy.

Week 1: building the first reinforcement-learning MVP for bottle grasping and lifting with Unitree G1.

Final report: optimizing the inspection pipeline and completing the PoC with Go2 and G1 camera compatibility.

Week 3: replacing unstable VLM reasoning with grounded detection and deterministic geometric risk policies.

Week 2: connecting the Go2 camera to a persistent local VLM pipeline for continuous visual inspection.

Week 1: designing and implementing the first end-to-end MVP for a robot visual inspection agent.

Jun 1, 2026

The first post, but no technical content, just a reflection.