How to choose your hive scale?

The last few years have seen the emergence of a large number of products for hive monitoring. Specifically, the offer of hive scales has been greatly expanded. In France alone, the ITSAP Institute lists about 20 models. At the global level, there are about 40 commercial products and about thirty open source products.

Now, in front of this diversity, how to choose your hive scale? The available solutions are all different but they can be classified in 4+1 criteria.

1. Hive scale architecture
2. System and transmission architecture
3. Software features and daily use
4. Support and durability over time
5. Cost

So we'll break down each of these points. But before we go any further, you should know that there is no "winning" solution. There are only solutions that fit needs. Of course the need is always to monitor hive weight remotely, but we will see that this is not enough as a directive: You must know what to monitor, under which conditions and for which decision-making.

So let's move ahead visiting all of the topics listed above. We start with scale architecture.

Hive scale architecture

The scale architecture will be an essential factor in your choice. Despite the diversity of the offer, the technical choices for weighing a hive are not infinite. In practice, and except for some marginal exceptions, they are reduced to 3 architectures which are characterized by the number of sensors (or load cells) installed.

Single load cell scales

This is the architecture of the mythical Capaz scales. In this post I will not refer to any manufacturer, but I think that Capaz Gmbh can be mentioned for two reasons: The first is that Mr. Guido Koch from Oberkirch in Germany was the pioneer of the beehive scale about 15 years ago. The second one is that as this brand has withdrawn from the market, we can talk about it as one talks about the great stars once they are gone.

So let me pick up the thread again: We were talking about scales with a single load cell. This one is often located in the center of a double H-shaped metal structure. This solution is certainly one of the most accurate and you will find it in your butcher's shop or in any store that needs legal weight metrology.

While acuracy is the asset of this architecture, there is also a major drawback which is the scale weight (the butcher doesn't care, you do if you're migrating your hives often). Count 10 to 12 kg per scale. The heavier it is, the more accurate it will be because the goal is to have a super rigid structure around this single load cell which must be the only deformable element (to measure the strain and then the weight).

This architecture has a second disadvantage which is the stability of the hive on the scale. As all forces pass through a single path which is the load cell connecting the upper pan with the lower base, it is by definition not supported on more than one point. The butcher doesn't care because he only weights slices of ham or ribs, but if you set on it a beehive with 3 supers it will swing a little. The stiffer the scale, the less this oscillation will be visible, but the heavier the scale will be (to return to the previous point).

Typically in this category, if the load cell is of good quality (require C3 minimum according to OIML classification ) the accuracy can go down to +/-50 gr. This architecture can be recommended for scientific or research use when high precision is an essential requirement.

What about hive migration you may say? Well, these models have been widely used for migratory beekeeping in recent years (first models on the market) and indeed they have the great advantage of being insensitive to leveling and installation. However, their weight and size are an operational constraint. It is possible to bring 12kg more to the apiary to have a high precision measurement, but experience shows that it is better to bring 2 scales to the apiary than one, because it is much more useful to have a (slightly less accurate) weighing on a few hives than the most accurate weighing of a single hive.

Two load cell scales.

Scales with two load cells often look like a bar. This bar, which is placed at the front or back of the hive, has a cell at each end.

This device weights half the hive and does weight_hive = weight_scale x 2 assuming that the center of gravity is fully centered front/rear.

The advantage of this architecture is its lightness and compactness. Around 1 kg for this scale, which makes it very easy to handle and install. The cost will also benefit from this material saving as well as the transport to ship it to you. The disadvantage is that the hive must be properly seated and leveled. You will need a similar sized wooden bar on the opposite side of the hive to make sure that the hive is well supported on all sides.

Typically, this category of scales is adapted to sedentary beekeeping. When the bases are well in place and we can ensure that they are levelled, they will give excellent results with an accuracy of around +/- 100 to 150gr depending on the model. It is certainly one of the flagship products for hobby beekeepers.

Four load cell scales.

In this category we come back to the scales that perform a complete weighing of the hive. The 4-load-cells architecture combines the advantages of the two previous categories. This means that the complete weight of the hive is measured (no centering or leveling constraint) with a guaranteed stability (the structure is hyperstatic). Those scales weight around 2-3kg due to an economy of materials in the structure.

 

This scale architecture is ideal for migratory beekeeping because they can be used in severe working conditions (night work) without the need to spend too much time on installation. Even on sloping ground and with a wobbly base the readings will be of high quality.

This architecture is also a good candidate for pallet integration.

In terms of accuracy, it will be between the two previous categories. Count on average +/-100g. But it will depend on the design and the grade of the load cells (we find C3, C2 and without classification). As for the size, it can join one of the two previous architectures. Either it is built in tray mode with a rigid structure that holds the 4 load cells, or it is built in double bar mode.

The following table summarizes what we have just seen:

 

Architecture	Advantages	Drawback	Usage
Single loadcell scales	Full hive-weight measure Accuracy ~50g Insensitive to leveling and seating error	Footprint Weight ~12kg	Scientific and research
Double loadcell scales	Lightweight < 1kg Small footprint Easy to install Cost	Half hive-weight Accuracy 100-200gr Take care of the leveling and the seat	Sedentary beekeeping
Four loadcell scales	Full hive-weight measure Lightweight < 3kg Small footprint Easy to install Unaffected by leveling and seating defects Palletizable.	Depending on the model: size or cables	Migratory beekeeping   Sedentary beekeeping

Now that we have seen the hive scale hardware architecture, let's move to their system communication architecture

System architecture

Connected devices rely on access to a network to transmit their data. In this context we will have two types of products depending on whether they work alone or in a network:

• Autonomous communicating sensors: They embed the communication system to the cloud. In this case the product is completely autonomous, with its own subscription.
• Sensors operating in a local network: typically the sensor communicates in a local network to a Hub or Gateway placed in the apiary which mutualizes the means to communicate to the cloud. This networked solution allows to operate a larger number of sensors (scales, brood probes, weather stations, bee counters, cameras etc.) with only one subscription which is the one of the Hub. Or even without subscription if the local communication protocol is Bluetooth and you decide to perform manual synchronizations with your smartphone.

System architecture	Advantage	Drawback
Autonomous sensors	Autonomous device stand alone	Hardly expandable One subscription per device
Sensors in local network to a gateway	Monitor several hives and/or apiary properties with a single subscription Shared transmission function. Manual synchronization possible	Device associated with a Hub for transmission to the cloud

Transmission network

Then we can distinguish the communication networks used by the sensor. In this register there are two main categories:

• Those that work with cellular networks (2G, 3G or 4G). They include a SIM card (often multi-carrier) that can connect to one of the general networks available.
• Those relying on LPWAN networks dedicated to connected objects such as Sigfox or LoRA.

Transmission system	Advantage	Inconvenient
Cellular Network 3G/4G	Access to multiple carriers with a single SIM Network coverage test with your smartphone. Rely on the generalized network (general public)	Higher energy costs.
LPWAN networks (SigFox, LoRa)	Low power consumption	Associated with a single network and its availability. Difficulty in assessing coverage without a dedicated device. Limited data bandwidth

Each network has its advantages and disadvantages. Cellular networks consume more energy than LPWANs, although in recent years they have greatly improved their performance. With LPWAN networks there is less power consumption but the user is bound to a single carrier. On the other hand, the user of cellular networks is more likely to be covered by one of the available mobile networks. Note that domain is extremely dynamic and we are seeing improvements and game changes (read the Sigfox story) over the last years. Hard to tell which will be winners in the near future.

Cellular networks have the convenience that you can assess coverage directly with your phone (at least with your own carrier) while LPWAN networks require a dedicated device to test the network.

These networks all have a relatively high coverage (>85%) of the territory. The concern is that beekeepers love to hide in the remaining 15%! So the best you can do is to go and check on each of your locations that you reach the expected connectivity.

Other practical aspects

Beyond the above considerations, there are some practical aspects that can also make it easier to use the equipment on a daily basis and that should also be considered. Here is a non-exhaustive list:

• If the devices have an internal memory, to save the readings and to be able to recover them in case of communication interruption
• The battery life of each device is an important point : 1, 2, 4 years or lifetime?
• The nature of the batteries: are they market standard or a specific model?
• The possibility to operate without remote antennas or other wires.
• Can you connect them with your smartphone for a face-to-face or in field check-up?

Software Features

So far we have considered the hardware side. The devices and their transmission. However these devices are not an end in themselves, they are a means to achieve our goal of monitoring the hives remotely. So what really matters is to determine how efficiently we can use the information they provide.

On this point let's consider several aspects:

• Availability of raw measurements
• Availability of automatic alerts
• Availability of advanced analysis and algorithms
• Availability of related data (weather, annotations, hive sharing, etc.)
• Traceability of each sensor's location over time (which hive/apiary)
• Mobile and off-network availability
• Collaborative features

The raw measurements

The basic service is to get raw measurements: the hourly weight, sometimes the ambient temperature or the brood temperature are quite common measurements. On these measurements, some platforms offer automatic alerts (overweight for example)

A weather service

In some cases you also get a weather service. Maybe even with forecasts, 3 days or 10 days ahead. This is far from being a minor detail. Having an integrated package is what makes it easier to make causal links and taking decisions.

Ensure data traceability

Another important aspect is to ensure that the data can always be used. Over the course of a season, you will move your scale from Apiary A to Apiary B, from Hive 1 to Hive 2. The scale will measure the weight continuously, but how do you sort it out at the end of the season? This is where location tracking comes in. If the system allows you to define the various locations of your scale you will be able to come back to it long after. Next year, for example, to find out how the honey flow was on Apiary A in 2020.

It is at this level that the monitoring system becomes a major asset for any beekeeper. It allows him/her to understand and compare from one year to another, from one apiary to another, from one hive to another. On other industries devices can seat in the same place for years. No need for traceability in this case. But for beekeeping, traceability is an essential feature to reach data capitalization.

If you don't do this association work, you will use the system in its basic function, as you go along. It will be a decision support tool in the moment, but nothing more. You will have missed a much more powerful use, which is to capture the data over the years to help you make better decisions based on your (data driven) past experiences.

Algorithms that speak to the beekeeper

Raw measurements can be processed with algorithms to provide beekeeper information such as weight Gain (variation over a given period) or Productivity (variation restricted to the work of the bees only) or the amount of brood in the hive and its development status. These algorithms bring a real advantage to the beekeeper who thus has access to methodically digested information, which speaks in the jargon of his profession.

Developing these tools is a full time job. It requires time and skills. That's why these tools are not yet widespread. Expect to pay a bit more for them.

Mobile app and desktop app

Having a mobile application allowing information access at any time (in masked time, in your roadway between two apiaries) with or without a network connectivity is a real advantage.

The web application, on a large screen, will be used to perform analyses over longer periods. By manipulating data and time horizons more easily. On the other hand the mobile app is the tool for action times. You will use both.

Collaborative features:

In some use-cases you will have multiple users like colleagues or fellows for example. A practical feature is to have the app installed on several smartphones fully synchronized. If your colleague has moved a scale, or added a note, you will automatically find it on your smartphone.

Hive sharing between two different accounts is also a feature that allows you to give access to a colleague who works in the same area and with whom you share your measuring equipment.

Here is a summary:

Software features	Example
Raw data	Hourly Weight and temperature of the hive(s)
Automatic alerts	Weight threshold, overheating, weather alerts, brood threshold, hives in trouble.
Advanced algorithms, beekeeping jargon.	Weight gain Productivity Brood volume Egg laying recovery/stop Risk of swarming Forecasts
Access to related data	Local weather Annotations, Blooms Nectar-flow conditions
Location traceability	Position over time of the sensor (hive/hive house)
Mobile and off-network availability	Mobile app on Android/iOs Access to off-network info On-site troubleshooting Web platform
Collaborative features	App on multiple smartphones automatically synchronized Share apiaries between fellow accounts.

Support and sustainability over time

The sanitary and climatic context that we live in pushes us to manage our hives much more finely than our elders could do it. Adopting a hive monitoring system is an important decision. It is to engage in a necessarily different practice of beekeeping. It transforms your practice. The beekeeper who commits to the use of a connected scale must know that nothing will be like before. He/she will be also committing to a new learning process shedings light on what used to be beliefs or suppositions and drastically reinforcing the decision making.

In this context, the choice of the system he will adopt is a structuring choice. Because the risk is that after a while, he has acquired several equipment from various suppliers and that the spread of data and information does not render any service in the long term. (read How to choose an apiary monitoring solution ).

The cost of the system

Here is the topic I left for last. It is a kind of counterbalance that I am doing here because in my experience the cost of a system is perhaps the first thing that everyone looks at when making their choice.

If this is your first step into the field of hive monitoring, remember that you are not buying a sweater, not even a "scale". Above all, you are making a commitment to yourself and to a manufacturer for a period of time that you both would like to be as long as possible.

The final cost is nothing but the result of all the previous considerations. I personally like the saying "You always get what you pay for".

The last word

If you've read this far, congratulations, I hope you've learned something and have acquired some guidelines for your choices.

Don't hesitate to share your comments below!