By James Wilkes
Hive Tracks is beekeeping software created by beekeepers for beekeepers with the goal of improving the quality and experience of beekeeping for everyone. It is a web application, which simply means you can access the Hive Tracks software through a web address, hivetracks.com, using any internet enabled device including smart phones and tablets. The vision for Hive Tracks was born in the minds of two beekeepers who live and keep their bees in the Blue Ridge mountains of North Carolina, an area rich in beekeeping tradition and well known for tasty honey varietals including world famous sourwood honey. These two beekeepers, Mark Henson (above right) and me, James Wilkes (above left), dreamed of utilizing cutting edge technology to build easy to use tools and services to help beekeepers, ourselves included, have healthy and productive honey bee colonies. Our hope is that by maintaining information like records of inspections and events in hives and beeyards, every beekeeper will be equipped with the information needed to make wise management decisions for their bees. Whether you have a couple of hives in your backyard or a couple of hundred in varietal honey production or several thousand colonies for pollination, knowing the current state of your bees is essential to being a successful beekeeper. This article gives a brief backstory of the development of Hive Tracks.
The story of Hive Tracks is full of coincidence or divine providence or whatever you want to identify as the cause of events that come together in just the right way. One example is how the creators of Hive Tracks found each other.
Mark Henson is a professional software engineer with 30 years experience in software development and a masters degree in computer science. He lives in Boone, NC, with his wife and daughter and telecommutes to work very early in the morning with a software team in Great Britain. He has been a backyard beekeeper or eight years with a hive count varying from a few to more than 10.
I am a Professor and Chair of Computer Science at Appalachian State University in Boone, NC, where I have been teaching for 23 years and I have a PhD in computer science. My wife and eight children and I live in Creston, NC, which is very close to Boone, on Faith Mountain Farm. We are sideliner beekeepers as part of our farm business with about a 100 hives, although we had about 40 hives when Hive Tracks was born.
Mark and I moved in the same beekeeping circles, namely the Watauga County Beekeepers Club and as honey sellers at the local farmers market, so it was no surprise when we were considering our nascent ideas for Hive Tracks that mutual acquaintances suggested we get together. So, on a snowy afternoon in late February 2009, Hive Tracks was born over lunch.
From the beginning, our ideas were very similar. I was standing at a hive in my beeyard that previous summer of 2008 ready to perform an inspection. I scratched my head trying to remember what I observed the last time I was inspecting this hive. Ever done that? In a moment of clarity, I caught a glimpse of what it could be like – walking up to a hive, a handheld mobile device (smart phones were not so smart back then) recognizes the hive being inspected and shares information with the beekeeper that will help with this inspection, like the health or strength of the hive, the queen status including her age, any unusual observations at the last inspection, medications or feed that should be checked, honey flows in the region, tips on what to look for at this time of year, etc. I saw the future, but did not see how to make it a reality.
Mark’s innovation came to life in a conversation with his wife on a long car trip home during Thanksgiving of 2008. With two years of beekeeping under his belt, Mark’s interest in improving his own beekeeping combined with his software expertise resulted in him being driven to create a prototype hive information system being in place by Christmas of 2008, a month after his initial brainstorm. It was this prototype that he showed to me at our first meeting, bringing life to ideas that previously lived only in my head.
Mark had already shown the prototype to Shane Gebauer of Brushy Mountain Bee Farm at a bee meeting and later we showed it to David Tarpy from the Entomology Department at NCSU at the North Carolina state bee meeting. Shane and David have both encouraged us and have been supportive of Hive Tracks over the years.
Ideas are relatively easy to dream up, but implementing them is the real challenge. In the case of Hive Tracks, a plan was made to develop the first production version of the software with a launch date of August 1, 2010 coinciding with the 2010 Eastern Apicultural Society Conference, which happened to be held in Boone, NC that year. One of those happy coincidences referenced earlier. The first production version of Hive Tracks was created during the year preceding the conference with innovative features including a digital representation of each hive in an apiary based on the hardware components of the hive and graphical indicators of hive health and queen status. The first of its kind (that we know of) hive editor allows the beekeeper to maintain the proper hive configuration as it changes throughout the season. Mark never knew how many different components beekeepers used until trying to represent them all! Components are still being added to this day so send us your favorite non-standard component to add to the list. Just kidding! Following the initial launch, we were excited to see 400+ users signed up by the end of August.
By the end of 2010, 800+ people had registered accounts with Hive Tracks with no real marketing other than a favorable review in Bee Culture and word of mouth through bee club presentations. Growth in user accounts has always been steady with over 6000 users by the beginning of 2013 and continuing to this day with our current list of more than 13,000 users. Most of this growth has predictably been in the United States and more specifically in the eastern half of the U.S. where there are more backyard beekeepers who make up the majority of our users. The average number of hives per user is around five, but we have a surprising number of sideliner and even commercial beekeepers who are finding it useful for their operations (more on this aspect later!).
A fascinating aspect of our growth has been the number of Hive Tracks users around the world; a great illustration of the interconnectedness of our world via the internet. We now have users in over 130 countries, with the top numbers being in United Kingdom, Canada, Belgium, Greece, Australia, and New Zealand. Numerous users have offered to translate Hive Tracks to their native language, which we expect will happen in time. And then there is dealing with the variety of beekeeping practices around the world including different hardware types, queen breeds, treatment strategies and products, and even the reversed calendars of the northern and southern hemisphere. So, although there is great commonality among beekeepers worldwide in their love for bees and the desire to keep them well, providing software that is relevant to them all without making it unwieldy is a challenge! However, it is a challenge we are facing head on by developing a strategy for engaging more fully in regional beekeeping markets by creating “localized” versions of Hive Tracks built specifically for local beekeeping practices.
The example in which we are most advanced in working out this plan is in Tanzania. To get us to Tanzania, I must relate to you another installment of a “what a coincidence!” story. I travelled to Kiev, Ukraine in October of 2013 for the Apimondia 2013 conference representing Hive Tracks with the simple goal of gaining a better understanding of the international beekeeping market. Before I left on the trip, a business friend suggested that opportunities tend to come from unexpected places and to be on the lookout. So, it should not have surprised me when I looked across the big vendor hall with thousands of people milling about and see a familiar face of a lady I had only seen through facebook on her Follow the Honey business page. I approached her and said “Are you Mary?” and she replied, “Yes, who are you?” After the introductions, she immediately whisked me over to the Tanzania vendor booth where I was warmly welcomed by the government contingent and David Camara, the owner of National Beekeeping Supplies, Ltd, from Tanzania. David and Mary and the Tanzanian government are working to market and promote honey from Tanzania. Hive Tracks was recognized as a useful tool to accomplish their goals by helping individual beekeepers be organized and more productive. We have partnered with David to make Hive Tracks a part of the beekeeping landscape in Tanzania and Africa.
Free Helps Growth
Hive Tracks is free (you can also pay, but more on that later!). How can it be free? We are generous and want to share this tool with everyone, but it is not that we are independently wealthy and do not need to make money. No, we made an early business decision that we would always provide a free version to make it as easy as possible for all beekeepers to have a great tool for record keeping and to fuel the growth of our user base. We may have set the bar too high on how much we are giving away (like I said, we are generous), but our strategy is in line with many other cloud based software providers like Google, Survey Monkey, and Mail Chimp to name a few. The thinking is that once you have a large user base, you can then figure out ways to monetize through advertising and offering premium services to the users who already find value in your free tool.
From early on we accepted donations, and a few users did donate which helped with some of the costs, but mostly expenses were out of our pockets along with a healthy dose of sweat equity (funny, that sounds a lot like beekeeping to me). About a year and a half ago, we added a business partner with the goal of advancing the development of Hive Tracks into a sustainable business. Toward that goal, in December of 2013 we released a completely rewritten version of Hive Tracks with better graphics, more features, and faster response.
Shortly thereafter in March of 2014, we released Hive Tracks Pro, a subscription based paid version with very modest price tag. After three and a half years of users telling us what new features they wanted and based on our own needs, the Pro version includes a number of value added features including additional hive types like nucs, uploading of photos and videos, a hive hardware inventory, integrated calendar and todo lists, group sharing of data, and recently an Android mobile app for offline work.
We also do some advertising and expect that to increase. We do plan to make money, but also want to make it accessible to as many beekeepers as possible. Trying to find that balance is the challenge. We hope people will see Hive Tracks as their “digital hive tool” and make it a necessary part of their beekeeping practice and will choose to support us.
Tip of the Iceberg
As the realization of the vision we had more than five years ago continues to take shape, we believe the fun has just begun in terms of what is possible with only the tip of the iceberg exposed so far. Hive Tracks is built on a solid foundation of fundamental data of yard, hive, and inspection information that is important to all beekeepers no matter the size of their operation.
Building on this foundation, much more is technologically possible including collecting data from instrumented hives (sensors for weight, temperature, humidity, and audio), connecting beekeepers through social media, forums, and information sharing, analyzing the combined data of beekeepers (big data for beekeepers), and an endless list of cool features including any number of reports like honey production per hive or per yard, longitudinal queen performance, hive success when started from a nuc vs. split vs. package, etc.
The area of instrumented hives has always been one of high interest to us since combining observational data from Hive Tracks with instrumented hive data yields incredibly valuable information for a beekeeper. Perhaps even more importantly, this combination of data from a world wide network of beekeepers would be a very strong data set for honey bee research. In the past three years, there has been rapid development of commercial solutions for the hive scale problem with multiple vendors in the market now with one of the most extensive monitoring systems being offered by Arnia (featured December ’14 and this issue in Bee Culture articles), which includes weight, temp, humidity, and sound, as well as analysis of the data to detect hive events.
The potential benefits to our beekeeping experience offered by technology are endless and exciting to consider and pursue, and I invite you to join your fellow beekeepers and us at Hive Tracks as we strive to be better beekeepers by equipping ourselves with the right tools and information to make wise hive management decisions. Hive Tracks is committed to pursuing excellence in beekeeping and will continue to work to add native features and functionality useful and affordable to all beekeepers. In upcoming articles, I will dive into more details of how Hive Tracks works and what value it can add to your beekeeping experience. In the end of course, the most important part of your beekeeping is still to be in the bees!
James keeps his bees in too many yards in the mountains and foothills of northwest North Carolina
By- Huw Evans
The Benefits of Electronic Data Collection
When compared with manual collection, electronic hive monitoring offers scientists a less intrusive means of collecting more objective data at much higher resolutions, several times per minute if required. Multiple parameters can be simultaneously measured and recorded. The measurements are more consistent as probe positions and settling times do not vary. Remote monitoring simplifies data management; the data is automatically collected and stored in a database saving a huge amount of work while minimizing any potential ‘mix ups’. The data can be graphed from within or downloaded from a single user interface accessed from any Internet enabled device. Which measurements are taken and how often can be configured remotely from anywhere in the world with Internet connectivity. The same user interface allows beekeepers to enter inspection data such as hive manipulations, diseases, Varroa counts and treatments.
Bees do Nothing Invariably
One of the issues facing scientists researching bee health is that “bees do nothing invariably”, as we say in the UK or “bees, its like herding cats” as is more commonly used in the U.S. That’s certainly the case with my bees. On good years I think I’m a great beekeeper, on bad years I think I must have done something horribly wrong. The truth is despite doing the same thing every year I get very different results. OK, I’m not saying that there is not a correlation between how well you look after your bees and how well the bees do, but the health of my bees is a multi factorial thing and I’m only one of those factors. Our suggested approach to this problem is ‘Big Data’ supplied by thousands of monitors. Not only does the sheer amount of data provide buffers to variance, (or buffers to deviations encountered on smaller scale data samples) it can also reveal trends when correlating bee health geographically with potential environmental stressors and different beekeeping practices.
Agricultural practices, be it the use of Plant Protection Products (PPPs) or the landscape picture being dominated by vast monocultures, have been implicated as one of the major causes of pollinator decline. Effects of different PPPs, presence of wildflower margins or the proximity to wooded area could all be considered statistically on a wide scale. Similarly, correlations could be drawn with different habitats, urban or rural, managed and unmanaged. The data could also be used to identify trends between bee health and the proximity of things like electric power lines, mobile phone masts, wind farms, electric trains or even motorways.
A recent study at Southampton University has shown that diesel exhaust, in particularly nitrous oxide, rapidly degrades floral odors used by honey bees for the recognition of floral nectar sources such as Canola (Oil Seed Rape). This could have a negative impact upon a honey bee’s foraging efficiency and the pollination services that the bees provide.
Beekeepers ultimately have bees’ interest at heart but making correct management decisions is not always straightforward. There are no hard and fast rules, there are recommendations but experience shows quite different results. Varroa treatment is an obvious example, do different treatments disrupt normal colony behaviour and if so how badly?
What effect does breaking the brood cycle have on long term survival? Does queen age affect colony’s future prospects? What type and amount of supplemental feed should the beekeeper provide? There is plenty of circumstantial evidence but surely a large mass of data collected over long periods of time over a wide geographical area will identify real trends and help clarify these and many other bee husbandry issues.
Suitability of Bee Breed
One species of honey bee, Apis mellifera, has evolved to survive in widely varied environments and climates, which has through the course of evolution led to the emergence of a number of locally adapted subspecies. However, man has moved bees about the planet without giving much thought to the consequences. Most of us believe that local bees do better than the imported bees, but still there are plenty of queens and packages of bees being shipped around the world. Others argue for greater genetic diversity. Trending the performance of different subspecies in different climatic conditions could provide the necessary empirical evidence to help better understand bee breed suitability or even to be used for legislative purposes.
In recent years there has been a lot of debate regarding factors effecting queen health and fitness. As shown in the previous article, instability in brood temperature data can reveal when a queen fails or stops laying. The benefits of being able to trend queen failure geographically are obvious as we would know with considerably better precision when, where and in what order queens were failing. However, in a similar but opposite way, monitor data can also show when queens start laying, which in turn gives us a pretty accurate idea when they mated. For most beekeepers this is currently a bit of a guess as we are trained to leave virgin queens undisturbed for several weeks to mate. Bad weather at the time of mating has been implicated as a potential cause of drone laying queens. Our system also records weather conditions that can be correlated with mating times, even retrospectively.
A myriad of hive types exist and new ones are being developed continuously using new materials and technologies which can effect, for example, thermal insulation and humidity. Wide scale mapping of, for instance, hive humidity levels against hive type in different climatic conditions may help recommendations to beekeepers for more efficient housing of their bees.
Daily flight profiles
The importance of taking measurements regularly and simultaneously should not be underestimated. Using acoustics, we can economically plot daily flight profiles on a scale and at resolutions that would be practically impossible by human observation as flight profiles vary widely throughout the day. An example of this is shown in fig 1 where we can see that following a flurry of flight activity in the morning there is a lull in activity around lunch time. There is then a second wave of flight activity in the afternoon, which could well consist of foragers, young bees learning to fly in groups or even bees preparing to accompany the queen on mating flights. The ‘character’ of the flight profile therefore changes throughout the season.
Statistical consideration of daily flight profiles could provide a valuable insight into colony behaviour and potentially identify quite subtle trends in relation to exogenous and endogenous factors. Not only could the profile be correlated with other parameters such as what temperature they get going in the morning or how willing they are to fly in light rain, but also how this behaviour may be influenced by other environmental factors such as exposure to PPP’s or in-hive treatments.
Furthermore, an observer moving from one hive to the next throughout the course of the day could mistake a moment of inactivity for an inactive colony; a colony strength assessment would be perverted in a similar way. Therefore, simultaneous data collection is the only way to objectively compare different scenarios.
Weight of the hive is a pretty clear indicator of that colony’s productivity, both in terms of colony size and stores. Typically, the slope of the weight increase is proportional to the abundance of and the distance to the forage source. This information becomes more valuable when combined with measurements such as flight activity and meteorological data, as seen in fig 2.
The amount of flying necessary to bring in a certain amount of nectar can be considered a measure of ‘forage efficiency’. Typically, flight activity is proportional to the net weight increase, however we can see that following morning rain, despite an increase in flight activity, there is little nectar brought in. Fanning activity can be added to indicate nectar processing, showing when it takes place and to what relative degree. Like the daily flight profile, this gives us further insight into bee behaviour and even potentially useful circadian rhythm information.
Moreover, tracking weight over longer periods also offers insights into the occurrence and availability of forage. Hive weights were used by Dr Wayne Esaias to draw up maps of flowering vegetation, which in turn have shown the effect of urban warming on the timing of flowering. When there is dearth, the resulting drop in weight is a useful indication of the colony’s metabolism, in other words, how much energy the colony requires when there are no external sources of nectar are available.
Electronic hive monitoring can also help explain sudden and unexpected colony loss. When a plane crashes, all flight data preceding the crash is saved on a flight recorder or ‘black box’. A lack of ‘black box’ data often makes it difficult for scientists to establish the cause of colony collapse. In the last article we demonstrated that by considering a sudden drop in weight with a sharp increase in flight activity successfully diagnosed robbing as the cause of a colony’s failure.
Recently a beekeeper was transporting bees a relatively short distance, on arrival the weather was bad so he returned 24hrs later to open the colonies and discovered one of the hives was full of dead bees. Luckily that colony was being monitored, so we had a look to see if we could work out what had gone wrong. From Fig 3 we can see the day started with a nice healthy colony with a stable brood temperature. There was then a sudden increase as the brood temperature rose sharply to 46°C. Initially we suspected some kind of hornet attack, however using acoustics we were able to include fanning activity at which point the cause became obvious.
It was a warm day and the bees could have done with a little bit more ventilation. As they began to warm up, they began to fan which generated more heat and this positive feedback cycle lead to self-destruction. From that point on the brood temperature follows ambient temperature with a small lag as the temperature sensor is inside the hive. Using the monitor data we could pinpoint both cause and time of death, without this data we would still be guessing what had gone wrong. This is indeed a tragic tale, however it did furnish us with a new ‘overheating when being transported’ alarm feature.
It is not the intention of this article to be prescriptive about how the data may be useful to a scientist. However, the benefits of wide scale data collection with its ability to deal with variability and trend bee health appear obvious.
When compared to taking measurements manually, electronic monitoring offers less intrusive and simultaneous data collection at much higher resolutions. Sceptics may fear that relatively uncontrolled data may lack ‘quality’. However, this is where we can turn to ‘crowd’ or ‘collective wisdom’.
This data has benefits beyond bee health, “If the world is our coal mine then bees are our canaries”. Bees can be considered an indicator species, with a forage range of 10km2 they make perfect bio sentinels. A lot of good science has already used bees’ data to help to map air pollution, toxins in the environment and even global warming.
So how do you persuade thousands of beekeepers to start strapping hive tech to their hives? You simply offer them a useful beekeeping tool! arnia’s monitoring system brings benefits directly to the beekeeper which both aids recruitment and maintains beekeeper participation in wide scale long term field trials.
Large scale data collection is underway in the U.S. under the auspices of the Bee Informed Partnership. Arnia is one of the registered suppliers of equipment for this project.
As with the application of electronic monitoring to bee husbandry, the true value ‘big bee data’ can bring to scientists will continue to evolve over time.
Huw Evans is the co-founder of arnia, a research and development company that designs and builds hive monitoring equipment. arnia hive monitors are currently for sale in the U.S., for more information contact firstname.lastname@example.org.