The original note, as written at the time
Imagine a bathroom or toilet that is cleaned by a film of nanobots. These disposable nanobots are made with only one thing in mind — to reunite with their affinity partners. A core of the affinity partners is the main piece. First the submissive nanobots are sprayed in a kind of film all over the floor of the dirty bathroom. The central core is dropped into the drain of the bathroom. The reaction is activated using a remote system.
Ongoing development
A bathroom cleaned by a film of nanobots built for one purpose only: to reunite with their affinity partners.
The submissive bots are sprayed as a film over the floor. The central core is dropped into the drain. A remote system activates the reaction, and the film migrates to the core, carrying the dirt with it and leaving through the drain.
The elegance is that the cleaning is not a scrubbing action at all. It is a directed migration, and the dirt is not removed by force but by being caught in a crowd that is all going the same way. That is much closer to how a biological system cleans itself than to how a machine does.
The disposability is the honest part. A film that goes down the drain cannot be recovered, so it must be cheap and it must be harmless — and harmless in a drain means biodegradable, which is at odds with a nanobot that must survive long enough to do its work. That tension is the whole engineering problem, and it is a good one, because it forces the material design to start from what happens after.
The remote activation is what makes it a product rather than a coating. You do not want the film reacting when it is sprayed; you want it inert on the surface and triggered when the room is empty.
The risk is the obvious one: a technology that puts small mobile objects into a building’s plumbing is a technology that will eventually block a trap somewhere. Any real version needs a filter and a service interval.
The best cleaning systems are the ones that remove the dirt rather than move it around the room.