Source: HIVETECH SOLUTIONS, LLC submitted to
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
Funding Source
Reporting Frequency
Accession No.
Grant No.
Project No.
Proposal No.
Multistate No.
Program Code
Project Start Date
Sep 1, 2017
Project End Date
Apr 30, 2018
Grant Year
Project Director
Drennan, K.
Recipient Organization
Performing Department
Non Technical Summary
Honeybees are critical to our economy, food security, and environmental health. The honeybee pollination industry adds more than $15 billion in value to agricultural crops each year. This industry supports an estimated 90-130 crops, the harvest which accounts for up to one-third of the U.S. diet. These valuable pollinators ensure that our diets are plentiful with fruits, nuts, and vegetables. Not only are honeybees are a critical link in our food supply that affect productivity and pricing, they are important product of our rural communities. Our goal is to reduce national honeybee losses and help stabilize our food supply by providing beekeepers and growers with a new tool and new information.The problem we are addressing is that honeybees and their keepers are facing devastating losses each year. Beekeepers in the United States lost 44% of their colonies in the last year alone. The honeybee industry desperately needs a way to mitigate these dramatic declines and growers need a stable, healthy supply to ensure their crop yields. One of the primary causes of honeybee decline is the Varroa Destructor, a mite that can quickly overtake a honeybee colony if left untreated. Our solution to this problem is provide beekeepers with a new tool that uses advances in small, low cost sensors to give beekeepers more accurate information on the severity of varroa mite infestations and other health conditions. With more accurate and timely infomation, beekeepers can act quickly to treat troubled hives. Our research and technology will provide additional benefits to the industry: savings in time and labor for inspections, improved management, the creation of a honeybee health database, and a new way for stakeholders to work toegther at a national level to address emerging issues.
Animal Health Component
Research Effort Categories

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
Goals / Objectives
Using small sensor technology, we place data collection inside the colony, providing an unprecedented look into what health looks like from the inside. Through hypothesis-driven experimentation, our project goal is to capture and analyze data on honeybee colonies with varying levels of mite infestation. This insight will not only provide beekeepers with a critical perspective on how to best manage their hives, but will enhance their ability to adhere to best practices. By providing this dimension of monitoring for every level of beekeeper, we will improve honeybee health and mitigate population losses. Our project objectives include addressing several key areas of national agriculture policy related to honeybee health: 1) Establishing a baseline for sensor detected metrics of honeybee health 2) Developing a predictive modeling approach to monitor managed honeybees 3) Assessing stresses associated with mite infestation 4) Developing a comprehensive record method to support beekeepers, 5) Curating and sharing knowledge through publications and outreach 6) Creation of public-private partnerships to improve communication and implementation of policy and initatives.
Project Methods
In Phase I of the USDA SBIR program, we will develop baseline health metrics for honeybee colonies using embedded sensor technology. We will be able to track health through Varroa mite infection and treatment, and draw correlations of health between sensor measurements and behavioral evaluations. Using hypothesis-driven experiments, we will create a more comprehensive definition of honeybee health, as well as elucidate what mite infection and treatment looks like from inside the hive.We are currently developing a predictive control systems model as a baseline to understand how honeybees thermoregulate. Once we identify the factors that surviving colonies exhibit throughout the experiment, we will begin to parse out the monitor data for which factors are most predictive of health. We will analyze our data throughout this process using a generalized linear mixed model approach. Using this type of analysis will allow us to identify which measurable factors are critical for predicting metrics of honeybee health. We will use a model selection approach which will allow us to not only identify which factors predict metrics we identify as "healthy", but also identify which factors interact. We will evaluate the model with comparison to blind sensor data, visual inspection results, and continue to test against future data collection.To evaluate our technology, we will track performance and required maintenance internally as well as getting feedback from our intended users. For ease of use, will get feedback from beekeepers about current limitations and opportunities they see in our hardware design. We will log any noticable impacts made by the honeybees. Working with beekeepers and documenting interactions with the colonies, we will continue to prototype and test our technology in the field.

Progress 09/01/17 to 04/30/18

Target Audience:Our targeted audience for this project was commercial beekeepers and other researchers in honeybee biology. Changes/Problems:Varroa Destructor - New Research on Migratory Pest-like Behavior Varroa mites continue to be at the foundation of problems for honeybees. Our initial Phase I proposal focused on identifying Varroa mite infestation using a sensor-based honeybee hive monitoring system. In our conversations with commercial beekeepers, many are proactively treating for mites 4-6 times during the calendar year. Specifically, several miticide treatments are currently being used extensively without significant reduction in mite load before overwintering (DeGrandi-Hoffman et al. 2017, Greatti et al. 1992). Therefore, an early alert system would prove futile because every beekeeper should assume they have Varroa and should treat, especially through the fall. This is because Varroa mite populations grow exponentially from August until October during the colony's last brood cycle (DeGrandi-Hoffman et al. 2016). Varroa mites reproduce in broodcells. During the fall, honeybee colonies reduce brood production (Winston 1991), causing mites to grasp onto adult workers. With increasingly warmer fall temperatures bees are foraging for longer periods. Mites are being transmitted between colonies as bees fly and rob honey from nearby colonies as late as December in some geographic regions (DeGrandi-Hoffman et al. 2017). This has effectively made the Varroa mite a migratory pest. As such, an area-wide treatment approach should be undertaken to control Varroa mites. After extensive research and communication with experts in the field, we have developed such an approach that is more effective for managing Varroa mites while also increasing overall colony health and survival. In our initial Phase I proposal, we initially set out to test whether our monitoring system could detect Varroa mite infestation earlier. Three months into Phase I we determined that even if our hive monitoring system could detect anomalies caused by mite infestation, we could provide more value to beekeepers if we could design an actionable plan that leads to loss reduction. During our research into the characteristics of mite infestation, Dr. Gloria DeGrandi-Hoffman offered her expert insight into this problem, which she has been working to solve since the massive colony losses due to colony collapse disorder in 2007. Her extensive work on the Varroa mite as a migratory pest introduced our team to the practice of controlled overwintering as a potential method to reduce mite infestation and improve overwinter survival rates. With the approval of our program manager, Dr. Nowierski, we then decided to switch gears and test whether controlled overwintering could be the solution to many honeybee problems. Paired with our hive monitoring technology, controlled overwintering can give large scale beekeepers the ability to effectively manage thousands of hives. By providing beekeepers with a centralized storage option and remote inspection, we reduce the seasonal labor and transport of honeybees used for pollination services. Additional economies of scale are achieved during the seasonal distribution of honeybees for pollination to growers. In addition to health status, our technology can provide growers with more accurate inventories of honeybee colonies. What opportunities for training and professional development has the project provided?Theproject provided trainig and professional development for theproject team and several members of the beekeeing community. We worked with the Colorado Department of Agriculture to train apiary inspectors and provide education to commercial beekeepers. How have the results been disseminated to communities of interest?The results have been reported to the Colorado Department of Agriculture and informally to the Colorado Professional Beekeepers Association. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

What was accomplished under these goals? Our Phase I proposal developed into specific objectives related to accomplished five main goals: 1. Develop a prototype and test a mobile indoor controlled-climate apiary (MICA). We built a working prototype of a mobile indoor climate-controlled apiary (MICA) that keeps the honeybees clustered all winter at the target temperature of 4.5-7?C. The data reported from the environmental sensors showed that the MICA maintained the interior conditions consistently for 5 months without failure or incident. 2. Explore an indoor overwintering method using MICA to reduce winter losses. We found that honeybee colonies that were kept in the MICA were 72% more likely to survive the winter than honeybee colonies that were left outside to overwinter. By using our sensor technology, of the 10 colonies that went into the MICA, we can see we have at least 8 colonies that are still alive. Of 12 colonies that were overwintered outdoors, only one is alive. Comparing the probabilities of survival, colonies inside are significantly more likely to survive than colonies outside (ChiSquared test, p <0.001). In this proof of concept experiment, we show that indoor overwintering in the MICA significantly reduces the chance of losing colonies compared to keeping them outside, and they emerge from winter stronger, more robust colonies, ready for pollination. 3. Monitor honeybee hives with embedded sensor technology to determine the potential for detecting changes in hive health. Our technology can show how effective honeybees are at maintaining brood temperature, which is critical for healthy development. Using sensor data taken at the same time as observational data, we can track how colonies colonies thermoregulate over time. We know observationally when these colonies have diseases, like European Foul Brood, and we can compare sensor data from a diseased colony to a similar colony in the same yard to evaluate how metrics like temperature, humidity, and sound differ. After optimizing the SVM to predict death, we will apply this algorithm to predicting stress and disease. Being able to translate and correlate observational data with empirical data opens up a new methodological approach that harmonizes the experience of beekeepers with scientific rigor. 4. Monitor honeybee hives to identify mite infection patterns in the data and identify optimal treatment windows and effectiveness. Through our research and communication with beekeepers, we no longer believe there is an ideal miticide treatment window. Between nectar flows and recommended temperature ranges for treatments commercial beekeepers are already treating up to 5 times a year during all of treatment windows available. While we are working on algorithms to identify colony-level physiological changes with mite infestation, we believe monitoring population growth, queen status, swarming status, other disease status, and food storage may be a better use of sensors than mite infestation. 5. Develop a uniform inspection form and procedure. We have designed a online visual inspection report to be used in the field with cellular connectivity. We are working to streamline inspections and utilize standards for colonies to "make grade" or suitability for pollination services. Creating a simple scoring system has also allowed us to effectively group colonies for analysis. We are working with Henry Graham from the Carl Hayden Bee Research Lab to corroborate our rating system with his method of evaluating colonies for making grade. We are also working with the Colorado Department of Agriculture on virus and bacterial analyses to correlate with our observational data and will be getting this data in the Spring of 2018.