What we learn building robots for ports and hospitals: design, training, orchestration, engineering and the landscape around it.
The restroom cleaning category read vendor by vendor: what the incumbent machine actually does, what a Singapore restroom cleaner costs by statute, the only published rental prices in the market, why a camera is the binding constraint in this one room, and what we concluded before starting C1.
Twenty US humanoid, brain and fleet players read company by company: what each one ships, what it runs on, what it has actually deployed, and what its buyers say is still missing.
Twenty-eight Chinese robot and brain companies read one by one, with the macro numbers, the verified productive deployments, what overseas buyers say about buying Chinese robots, and where the gaps are.
What multi-robot systems actually share today, how embodied memory is built in the literature, what a Jetson Thor really runs, and the supply-chain rules a Singapore buyer works under.
Actuators, hands, tactile sensing, lidar, batteries, wheeled bases, structure, the rare-earth regime and what every incumbent's body does badly, with prices as seen in August and September 2026.
A deep dive on twenty US and twenty-eight Chinese robotics players, the state of multi-robot memory, on-robot compute and the parts supply chain, and what it says about where a Singapore full-stack team can do better.
The thesis behind Tidewell, tested against about 175 sources: the brain is the future, but not the way most people say, and Singapore's edge is specific rather than general.
First rolling chassis, one arm mounted, base narrowed for hospital lifts.