I received this note, and since I suspect the topic is of general interest, I am
sending a copy of my response to the list.

(referring to the wireless, remote reading, computerised hive weight monitoring
system)
> I presume a one-off of these  would be rather expensive. If (a developer)
> manages to get it right what do you think are the market possibilities?

Simply put, and not to exaggerate, I think that such a system would immediately
become a necessity for most commercial beekeepers in developed countries.

I spend over $100,000 (US funds) annually for fuel and labour.  If the savings
in fuel and labour were 10% -- a very low estimate, I believe -- that is
$US10,000 a year and $US3.33 per bee hive.

If the savings were 25%, which is not unrealistic IMO, then that is
$US8.33/hive/year, or $US333 per yard of 40 hives per year.  Splitting that in
some ratio between the beekeeper and the service would make both quite happy, I
would think.

Add to the above savings the increased income that comes from knowing what areas
are paying off and which are not, the savings in insurance and losses from
having real time theft monitoring (and tracking), the ability to monitor disease
and queen performance on a hive-by-hive basis and we are talking a real income
improvement for beekeepers using the system.  After all, in business, the net
profit is only a small fraction of gross receipts.  Even a 10% improvement in
revenue vs. expense can double the money in the pocket of the owner at the end
of the year.  Moreover the fact that, using the system we envision, hive visits
can be managed according to measured need, not guesses.  This can eliminate many
wasted days and, allow more leisure time or time to deal with other matters.

Such a system could be sold outright, but should -- IMO -- preferably be sold as
an ongoing service with wireless connections, maintenance, software, and regular
hardware and software upgrades included.  The beekeeper would be responsible for
all the field installations and report design (non-programming), the supplier
for all the rest.

Such a system would have to be entirely wireless, rugged as a hockey puck, and
probably solar powered.  If individual transducers were used, they would have to
be indestructible, wireless and non-obtrusive.  Maybe they could be in little
rubber chunks that somehow go under a hive or the corner of a pallet and are
polled by a wireless monitor/repeater in each yard.

Alternately -- and preferably, perhaps -- a rotating laser head in each yard
could monitor tiny changes in the height of a black mark on each hive and deduce
compression of the hive base spring component, and then calculate weight
changes, assuming there were some sort of compliant, yet resilient hive base
component placed somehow under each hive.  In this latter method, rain would
cause swelling of wood and settling of earth under hives.  This would affect the
readings, but software should be able to filter such effects out, especially if
an empty reference hive were used.

We must remember that the data we require is all *relative* and more or less a
continuous in time, not absolute single observations; therefore calibration is
not very important, but consistency between transducers is.   Absolutes might be
inferred, but are not the goal.

Simply put, we are watching weight *changes* in time and deducing meaningful
events from the patterns that arise when such changes are expressed as rates and
compared to other simultaneous data and defined historical events.  We are _not_
trying to say what each hive weight actually is at any given moment.  Empirical
data could conceivably be acquired by conventional methods and used for
calibration periodically if absolute values are desired, but I am not very
interested in this aspect and suspect that it would be tougher to get right.

allen