The idea of fully instrumenting hives with general purpose industrial grade
sensors has two problems: too much money and too many wires.  If the job is
to be done for a reasonable cost and without excess cumbersome wiring, more
cleverness must be applied. For instance, though one could get by with but 3
load cells per hive, that's still too much money and too many wires.  If one
were to build a tiny, low power radio transmitter with a lithium cell to go
inside a stiff spring, in which the spring determined the carrier frequency,
and seal the whole assembly into a rubber bellows, there would be a fairly
reliable relationship between spring compression, proportional to weight,
and carrier frequency.  The receiver for this and other functions would not
need wires to get the message.  Right now I'm working on self contained
wireless temperature sensors, and they must be cheap.

I wish in no way to impugn the good work others are doing with more
conventional instrumentation.  They are pioneers who point the way for
dedicated and cheap custom data acquisition.  Consider that when mouses for
computers were first made they cost hundreds of dollars each.

Another pioneering effort is putting up with the bugs of new instruments
until they are perfected, and perfecting includes working down the price.
Many patient beekeepers may be needed for that part.

A third task (actually first) is what Allen has been doing, thinking about
needs.

This is an excellent discussion!

Bill Morong