BeeLab Hive monitoring project

[MuswellMetro]

Thing is, all the electronic hardware bits are already pretty easily available, as small add-on modules. Whether as the computer (Arduino and Raspberry Pi are probably the two with the biggest following) or sensors (temperature, humidity, light, load, time, orientation, GPS position, etc) or outputs (mini-LCD, usb, audio, relays, SD card, bluetooth, wifi and mobile phone are just the start). Sparkfun are just one source for many of these modules.
There are specific problems though.
An obvious one is bee-proofing and beek-proofing the hardware so that sensors don't get either propped over like an intruding slug in the beespace or accidentally destroyed during frame removal for inspections.
One tricky technical challenge is powering the thing. Most apiaries don't have mains electricity. High capacity batteries (for a reasonable run-time) are expensive and large. The obvious solution is only to wake the thing up intermittently - which presents its own interesting technical challenges, particularly when trying to synchronize several devices to talk to each other while they are all awake simultaneously.
But where the real spadework lies is in the software to stitch the whole lot together, to read the sensors, format the data and present it. And naturally, a different, say temperature sensor chip, requires a different bit of software to read it. Any flexibility in the hardware configuration increases the software complexity, as all options must be covered and the actual configuration 'discovered' or somehow input by the installer.

Such projects, just like Government "IT" projects, need very very very careful definition and 'scoping' if they are to deliver anything before the money runs out.
How much has the BBKA put in?

This it is not sponsred by the BBKA just supported with "knowledge" hence why so far it is quite abysmal LOL,, too many cooks (Beeks),

The funding i think comes via the design Council to promote electronic to the community ( 200 units to be made) they present their progress reports to them at the V&A Museum https://designcouncilbeelab.eventbrite.co.uk/, http://www.horizon.ac.uk/Current-events/Bee-Lab-Breakfast/14801

the BBKA wants the data from the units publish on part of the blogs (software yet to be sent to me) and it has been suggest it would add to the knowledge base and predict things like how the flow develops across the country ( now can you see BBKA members doing that)

power as you said is the real problem, the beta design samples ever 60 sec and the AA batteries last about a week (expensive OUCH), early pre beta used D cells but the cheap battery holders would not hold solder ,so beta1 is AA batteries....old car battery comes to mind

 

[user 6228]

Thing is, all the electronic hardware bits are already pretty easily available, as small add-on modules. Whether as the computer (Arduino and Raspberry Pi are probably the two with the biggest following) or sensors (temperature, humidity, light, load, time, orientation, GPS position, etc) or outputs (mini-LCD, usb, audio, relays, SD card, bluetooth, wifi and mobile phone are just the start). Sparkfun are just one source for many of these modules.
There are specific problems though.
An obvious one is bee-proofing and beek-proofing the hardware so that sensors don't get either propped over like an intruding slug in the beespace or accidentally destroyed during frame removal for inspections.
One tricky technical challenge is powering the thing. Most apiaries don't have mains electricity. High capacity batteries (for a reasonable run-time) are expensive and large. The obvious solution is only to wake the thing up intermittently - which presents its own interesting technical challenges, particularly when trying to synchronise several devices to talk to each other while they are all awake simultaneously.
But where the real spadework lies is in the software to stitch the whole lot together, to read the sensors, format the data and present it. And naturally, a different, say temperature sensor chip, requires a different bit of software to read it. Any flexibility in the hardware configuration increases the software complexity, as all options must be covered and the actual configuration 'discovered' or somehow input by the installer.

Such projects, just like Government "IT" projects, need very very very careful definition and 'scoping' if they are to deliver anything before the money runs out.
How much has the BBKA put in?

my current solution is to embed the sensors into walls of the hive, but for that to work you need thick insulated walls . The computer is an industrial PC, and the software is the same as I wrote for my solar heating. You describe the sensor config to it in XML. The computer is 20m away from the hives, using 3.3v i2c which would be impossible but for some bus accelerator chips that increase the capacitance that can driven from 400 pf to 4000pf . At the moment I'm in the garden but it can see the 32 temp sensors in the hive using my iPad and the wifi connection as well as seeing if the shower water is hot enough the water and the rest of the solar heating system (42 different sensors).

 

[user 6228]

This it is not sponsred by the BBKA just supported with "knowledge" hence why so far it is quite abysmal LOL,, too many cooks (Beeks),

The funding i think comes via the design Council to promote electronic to the community ( 200 units to be made) they present their progress reports to them at the V&A Museum https://designcouncilbeelab.eventbrite.co.uk/, http://www.horizon.ac.uk/Current-events/Bee-Lab-Breakfast/14801

the BBKA wants the data from the units publish on part of the blogs (software yet to be sent to me) and it has been suggest it would add to the knowledge base and predict things like how the flow develops across the country ( now can you see BBKA members doing that)

power as you said is the real problem, the beta design samples ever 60 sec and the AA batteries last about a week (expensive OUCH), early pre beta used D cells but the cheap battery holders would not hold solder ,so beta1 is AA batteries....old car battery comes to mind

I would have thought it would be solar powered right from the beginning

 

[Hombre]

Hombre's opinions - on the sensoring of bee behaviour

For what it's worth: I believe that the place for temperature sensors are in the crown board. Set in so that they are essentialy flush with the under surface, but set in slightly and epoxied over to protect from over zealous use of the hive tool. Three or four sensors would permit interpolation of the cluster location in the winter months.

I have no real interest in the humidity - personal preference.

Temperature measurements I would take at between five and ten minute intervals. At once per minute the amount of redundant data will be a waste of both power and memory resource, with little in the way or unique data.

Weighing is a major problem. A strain guage or load cell must be protected from damage. A heavy hive placed directly onto a load cell will have an impact component that is dangerous. The hive must be free to move in the vertical plane unfettered. Given the variable weight of a standard hive, then is the actual weight of any real consequence? What is probably of interest is the change in weight from a given arbitrary minimum datum. In essence a 'nominal' empty hive weight of zero.
Continuous weighing would need to be part of a dedicated hive stand. Impediments would be straps, bricks, equipment swaps between hives and wind effects. Again the number of samples, around four an hour or so.

Temperature sensing could be both cheap and practical for all, but automatic weighing, remains to my mind, impractical except as a novelty, or for formal or informal personal research; due largely to the the mechanical constraints.

A lot of time spent in a nap state to preserve power. Local storage and once a day communications.

I believe that there will be a large number of similar projects in the next few years, and eventually there will be 'Aha!' moments that will turn out to be game changers. Low power communication has taken off seriously in the last five years and sensors are currently an area of great interest.

More strength to them.

 

[alanf]

power as you said is the real problem, the beta design samples ever 60 sec and the AA batteries last about a week (expensive OUCH), early pre beta used D cells but the cheap battery holders would not hold solder ,so beta1 is AA batteries....old car battery comes to mind

I like the idea of a "wired hive", I've even been thinking about how you'd put one together. My lines of thought started with Arduino, Raspberry Pi etc and there are examples out there http://openenergymonitor.org/emon/beemonitor they do sound better thought out that the "kit" although obviously other examples include no load cells.

The first design hurdle has to be that these boards are all designed for the lab, you really have to ruggedise them a bit and their power supplies before you could bolt the package to an outdoor hive stand. What you're saying about batteries reinforces that, the other example with a 12v car battery supply lasting a month or so as above seems a much more practical solution, or a pair of gel 12V lead acids that you swap every visit. On top of the basic Arduino or whatever, all the sensors and comms have to be added, and everything is custom built. You might want to improve the comms, maybe even include some anti theft tracking and GPS.

The thought occurred to me that there's a readily available processor with as much built in comms as you could wish, built in clock, GPS and dedicated battery casing in a rugged and compact case. It's a mobile phone. Tricky bit is that the inputs are all though the USB, but with something like item 111156136811 on ebay that might be an easier route than starting with a general purpose board. A new phone running Android 4 isn't cheap, but I have seen small tablets at under 50 quid. Or search for phones with damaged screens, they go on ebay for a few quid.

 

[Little John]

Weighing is a major problem. A strain guage or load cell must be protected from damage. A heavy hive placed directly onto a load cell will have an impact component that is dangerous. The hive must be free to move in the vertical plane unfettered.

Strikes me that weighing is more easily done by suspending the weight to be measured from the sensor(s).

So - base, couple of feet square with 3 vertical posts attached to it's perimeter. Affixed to the top of each post is a 32Kg rotary dial luggage scale (couple of quid each from Ebay, or your local 'pound' shop). Suspended from these luggage scales, and positioned just a couple of inches above the base is a platform onto which the hive is placed.

Each luggage scale is modified such that the dial rotates a potentiometer, rather than a needle.

Result - a cheap and robust weighing platform.

LJ

 

[itma]

...
The thought occurred to me that there's a readily available processor with as much built in comms as you could wish, built in clock, GPS and dedicated battery casing in a rugged and compact case. It's a mobile phone. Tricky bit is that the inputs are all though the USB, but with something like item 111156136811 on ebay that might be an easier route than starting with a general purpose board. A new phone running Android 4 isn't cheap, but I have seen small tablets at under 50 quid. Or search for phones with damaged screens, they go on ebay for a few quid.

And how often do you have to charge your smartphone?
However, various solar chargers and high capacity external batteries are readily available.
But generally connect via that usb port ...

But never fear, someone is bound to have been there already!
How about using an Arduino to just talk to all the sensors and communicate (both ways) with an Android phone - using Bluetooth?
You don't suppose anyone at MIT would have tried that, do you? http://www.amarino-toolkit.net

 

[user 6228]

A smartphone is overkill in money and power budget.
A simple 2g GPRS phone module costs £15 and has all the bandwidth you need. Correction the module is £10 but then you need to add an antenna and sim card holder. Standby current is a round 1mA but during Data transmission ~400mA.
2.4ghz radios for in apiary coms are £1.50 each. - these are from the individual hives to the phone module... range ~30m
There are processors that make ardinos look like an expensive power hog.

 

[m100]

There are processors that make ardinos look like an expensive power hog.

Power hog it is not.

ATmega328
Power-down mode WDT enabled, VCC = 3V 4.2μA
Power-save mode 32 kHz TOSC enabled 0.8 μA
Power Supply Current Idle 4 MHz, VCC = 3V 0.3 mA
Power Supply Current Active 4 MHz, VCC = 3V 1.7 mA

IMHO 'saving' any more power than that IS a pointless exercise.

 

[REDWOOD]

old car battery comes to mind
Why not make a small wind generator from an old printer servo motor and a plastic pipe

 

[alanf]

A smartphone is overkill in money and power budget.

A new working smartphone out of the box is entirely overkill. What's one with a broken screen worth? 20 quid? It's still a capable unix (sort of) box. The big power budget question is where it comes from. Access to a mains supply makes anything up to phone charging range trivial. Once you're isolated, either you're replacing half used batteries every visit (as per MM's description) or arranging for a periodic recharge. The advantage of using phone technology is the size of the market interested in charging from solar, wind or whatever.

And how often do you have to charge your smartphone? However, various solar chargers and high capacity external batteries are readily available. But generally connect via that usb port ...

Charging interval depends on what stays "awake" but a daily dose of current would be a reasonable requirement. Are there solar chargers that could keep it going over summer? Almost certainly, and you log the internal battery state to monitor how they are doing day by day. A USB OTG port will take input signals and charging current, according to spec.

But never fear, someone is bound to have been there already! How about using an Arduino to just talk to all the sensors and communicate (both ways) with an Android phone - using Bluetooth? You don't suppose anyone at MIT would have tried that, do you? http://www.amarino-toolkit.net

Everything is a variant on something that's been done before. Android and Arduino are both capable of doing something with a signal, would you need both? Admitted the options to read external sensors from Android are fewer, but as you pointed out yourself, software capabilities are also in the mix and you get a boost under that heading:

where the real spadework lies is in the software to stitch the whole lot together, to read the sensors, format the data and present it. And naturally, a different, say temperature sensor chip, requires a different bit of software to read it. Any flexibility in the hardware configuration increases the software complexity, as all options must be covered and the actual configuration 'discovered' or somehow input by the installer.

This route starts with a simple way of getting external sensor readings into a phone. Some want to start with a soldering iron, I'll start with a compiler. It's not holy writ (yet)

 

[pargyle]

Well ... I'm already halfway there and haven't spent a fortune yet:

But, my interest was in measuring the colony temperature and humidity in comparison to the external temperature for the purpose of evaluating the insulation of my hive and determining whether the colony (particularly over winter) is able to maintain a viable temperature.

Whilst it would be nice to have remote monitoring and a constant record retained in an SD module that could be downloaded to a PC I've settled for manual readings ie: 'moment in time'.

As the hive is in the garden I tend to take readings morning and evening when I'm working and perhaps another one or two during the day at weekends. I've been quite interested in the information that has already been recorded over the last few weeks and it suits my purpose admirably.

The colony temperature and humidity sensors sit in a small tube that is mounted in the crown board, the wires for the senders pass through a rubber bung in the top of the tube so there's no heat loss or draughts ... there is a mesh floor in the bottom of this tube so the bees can't get at the sensors and so far (presumably as there is no air flow) the bees have not propolised the mesh.

There are two further temperature senders - one exposed to the outside air and one located on the mesh floor inside the hive so I have four measurements:

Colony temperature (as at the top of the frames ... not in the cluster obviously but a good indication that they are working).
Colony humidity (At the same location as the temperature sender above)
Mesh floor temperature
Outside air temperature.

The actual monitors are contained in an Asda food safe box with a clipped seal tight lid and the wires passed through a hole that is sealed with a bit of blue tac ... all pretty Heath Robinson but it works ! The seal tight box is screwed to a hinged plate that is screwed to the hive stand - hinged so I can hinge it upwards to read the digital read outs without bending down and squinting !!

I may add a maximum/minimum thermometer for the external air temperature in due course before winter.

Having watched this thread I can also see the benefit of a hive weight measuring device but I'm struggling with that at the moment as my hive is very heavy and I haven't yet found a digital scale (ideally one that that I can weigh the hive by suspending it at each end) that will go high enough for the weight of my hive that costs less than a fiver !! Budgetary constraints ~ self imposed. Something will turn up !!

 

[alanf]

I can also see the benefit of a hive weight measuring device but I'm struggling with that at the moment as my hive is very heavy and I haven't yet found a digital scale (ideally one that that I can weigh the hive by suspending it at each end) that will go high enough for the weight of my hive that costs less than a fiver !!

Two possibles that have been seen:

Graham Royle did a talk that included photos of his set up of a few years back. An adapted floor included mechanical bathroom scales. As I recall there was some shrouding to protect from the weather and a flap with a mirror where he could read the dial once a week.

I did a simple lift with a digital luggage scale on each side of the base over a winter. Each side needed to bear up to 20 Kg to avoid taking the hive apart. For total weight, just add them together. There was somebody marketing an engineered version that included a hinged base and a levered pulley that registered on (yet another) digital luggage scale. The pulley and one sided lift meant the total could be estimated as 4 times the scales reading.

It's the difference week to week that in interesting; consistency of reading is more important than actual weight. A wider base might mean a reading from each end of a longer hive. Digital luggage scales (2 or 3 pounds online) tend to be limited to 40 Kg, a bridge and pulley could double the capacity to cope with a hive nearing 160 kg, but you're carrying quite a lot of kit by then. If you started with bathroom scales sized for 130-150 kg, levering onto scales at each end in turn could get you into the 250-300 kg range.

 

[Hombre]

Anyone able to re-purpose second hand Clear Blue processor chips and their limited display?

The Arduino power problem is with the USB power requirement. The Atmel chip itself can be made very frugal, as can the Picaxe series of chips.

 

[user 6228]

Anyone able to re-purpose second hand Clear Blue processor chips and their limited display? ...

you will need to explain a bit more what these are

 

[user 6228]

Yep, a very very long way to go, but it is a DIY kit to enable you to build on whatever you want (like mobile Data retrieval at extra Cost) so itreally only a starter kit for you to modify

The BBKA wanted Temp, Humidity and hive weight but somehow after talking to other beekeepers and several top bar owners the design team came up with a 20kg load cell,,hmmmmmm my 14x12 are over 50kg fully laden

so the designer should have been a beekeeper and they are not, they could not believe how heavy a 14x12 frame was

The DIY kit will be about £50-£60 and measure humidity, temperature and weight but you build it into your hive etc etc~~~ no commercial interest except if itest it i get one free, so test test

with a bit of thought the load cell only has to measure the change in weight. you can counterweight most of the hive weight out.

 

[pargyle]

you will need to explain a bit more what these are

Pregnancy test kits ...

 

[user 6228]

Pregnancy test kits ...

that clear blue... that trade mark is used in a number of markets

a bit like Triumph

 

[pargyle]

that clear blue... that trade mark is used in a number of markets

a bit like Triumph

Perhaps Hombre will come back and enlighten us ... fascinated to hear how he intends to recycle them ... if that's what he's talking about !!

 

[Hombre]

No, I have no magic to convert one or more of these, but I do note that there is more computing power in your average pregnancy tester than was used to land the lunar module on the moon. It was somewhat optimised for it's purpose though.

It does seem a shame that disposable micro controllers can be usefully recycled. It would be be an interesting concept of course.