Looking at the faults my espresso machine develops over time, I’ve often wondered if fault detection and error messages could be better. It’s probably safe to assume that the cost of including sensors for all possible faults and developing code to check and understand all those sensors is prohibitive and unreliable, so I’ve wondered about a different approach.

How about this. Stencil the chassis of the machine with a couple of layers of conductive paint in elaborate patterns, and then attach to that just enough electronics to send and receive electromagnetic pings through the paint, which would be distorted by thermal and mechanical stresses in the chassis. And then use AI to sort out the meaning and location of those distortions?

There’s already a company selling ‘electric paint’ for much simpler tasks. This one might need electronics closer to an SDR antenna than to a toy computer GPIO. Also more consistent fabrication, if that problem can’t be auto-calibrated away (along with variaton due to weather).

Why AI? Well, for the same reason as every other AI thing. Because I don’t know how to design it correctly but I can see that such could be designed in principle and AI could probably figure out how that and then some more. And, of course, variation in manufacturing implies a lot of autocalibration, which would make manual design even hairier.

Maybe this could be done in the cloud. Use some basic configuration to decide what’s normal, and if anything goes outside of that then have the machine phone home to say it feels a bit weird, and then have that service check the symptoms (aberrations in stencil response) against the symptoms witnessed in machines that subsequently got serviced.