Sensor technologies for enhanced beekeeping efficiency in South Africa: Insights from KwaZulu Natal
A blog by J. Ross (1), J. Godber (1), H. Sanders (2), C. Woodward (2), C. Kroukamp (3), A. Darker (4) & L. Alusala (4), Frontier Tech Implementing Partners.
This post captures learning from the first Sprint of our pilot — ‘Harnessing sensor technologies to improve beekeeping productivity in South Africa’.
(1) Crop Health and Protection (CHAP), 1.34 CHAP, Innovation Centre, Innovation Way, Heslington. YO10 5DG. United Kingdom.
(2) AgriSound, Entreprise Centre York St John University, Lord Mayor’s Walk, York. YO31 7EX. United Kingdom.
(3) CropImpi, 18 Neden Road, Montrose, Pietermaritzburg, South Africa
(4) UK Science & Innovation Network in South Africa, British High Commission, 255 Hill Street, Arcadia, Pretoria, South Africa.
*corresponding author: jenna.ross@chap-solutions.co.uk
Beekeeping contributes significantly to the South African economy through honey production, pollination services and enhanced natural capital, but challenges remain regarding sub-optimal pollination, biodiversity loss, productivity challenges and attracting new entrants to the sector especially from underserved groups.
This project addresses the aforementioned by assessing the use of internet-of-things (IoT) through in-hive sensor technology to deliver real-time decision support. The technology has been implemented across Europe, so the pilot explores the adaption and deployment in South Africa. Two provinces, namely KwaZulu Natal and Western Cape, were chosen for trial sites due to the presence of the African honeybee (Apis mellifera scutellate) and Capensis bee (Apis mellifera capensis), respectively, with data on optimal behaviours and hive conditions gather for the two different species and locations.
The project is a UK-South Africa collaboration between:
AgriSound, UK pollination management technologists,
Crop Health and Protection (CHAP), a UK Agri-Tech Innovation Centre,
CropImpi, a South Africa-based R&D organisation.
The pilot is supported by the Frontier Tech Hub, an initiative funded by UK aid from the Foreign, Commonwealth and Development Office (FCDO), and working closely with the Science and Innovation Network team in South Africa.
Details regarding the technology and outputs of Sprint 1 can be found here.
Sprint 2: Building on earlier work and expanding our reach
For Sprint 2, we investigated the following critical questions:
Will the 25 users in KwaZulu Natal value the benefits of data generated from the 100 sensors deployed in Sprint 1, and know how to respond to the data? This is important for understanding the customer needs, as well as gathering data for the African honeybee (Apis mellifera scutellate).
Can we identify and train a further 20 beekeepers from a wide variety of backgrounds from the Western Cape province, deploy device in their hives and record data. This is important for understanding whether there are any barriers to adoption for different locations and end-users, whilst also gathering data for the Capensis bee (Apis mellifera capensis).
Will the algorithm need to be modified depending on the bee species present in the region?
Can we generate data to inform a business case for entering the South African market? This is essential in developing the commercialisation strategy for tech partner, AgriSound.
Sprint 2: What we achieved
In sprint 2, data was gathered from the 100 deployed devices and the 25 users in KwaZulu Natal identified in Sprint 1. Results showed that 17% of devices deployed did not have sufficient signal 24 hours after installation and had to be redistributed. This is often difficult to check on-site, as the devices only upload data to the cloud every four hours to maximise battery life, and there is limited cell phone coverage and often stage 6 loadshedding. This means that the use of the devices may be limited in remote areas and during loadshedding. Of the remaining 83% of devices, data was successfully gathered. This demonstrated that the devices work when they have signal and work within the high humidity environment of KwaZulu Natal, which was previously unknown.
The 25 participants in KwaZulu Natal were also visited by CropImpi and progress questionnaires were completed.
Results showed that:
46% of participants said they had taken action as a result of the technology, after logging into the user interface and reviewing their hive analytics. The ability to monitor humidity and temperature data led some of the participants to move their hives or remove extra boxes for collecting honey. Swarming was also recorded and actioned. It should be noted that no actions were taken around theft alerts, as no hives were reported stolen or vandalised during the period.
29% of respondents were unsure about taking actions. The reasons noted for no actions included hives being based too remotely, hives being in resting period where no pollination activities were occurring and challenges/barriers around mobility or access to hives.
13% said it reduced time spent monitoring hives, 25% said it enables them to take quicker action on their hives.
No yield changes were detected due to missing the nectar flow – this will be reviewed again in Sprint 3.
63% of participants were satisfied and valued the technology. This result was linked to access to temperature data and swarming.
38% of commercial and hobbyist beekeepers indicated that the technology made life easier (e.g. able to access data on the overall activity of hives). Newer beekeepers were more focused on the fundamentals (e.g. catching swarms, predators and overall beekeeping practices).
In Sprint 2, we also selected 20 participants from the Western Cape to participate in the project. This was done to gather data on the Capensis bee (Apis mellifera capensis, as well as data on differing climatic conditions and demographics of beekeepers. However, 2 participants pulled out due to the delayed start date, leaving 18 participants in total. We aimed to capture a wide demographic of participants including different ages, genders, business sizes, literacy levels and income brackets to maximise learning around barriers to adoption (e.g. computer literacy), accessibility of the technology, as well as market demand.
Participant ages in the Western Cape ranged from 19 to 65 years old, there was a 69% male to 31% female gender split across the group, and around 6% of participants identified themselves as having a disability (Fig. 1).
Figure 1: Participant demographics in Western Cape (A) Age; (B) Gender; and (C) Disability
All 18 participants received an induction and training, covering topics including components and installation, hive health and tips for beekeeping. 100% of participants completed the quiz with an average score rate of 80% and 100% completed the interviews. An analysis was conducted to review the similarities and differences in users to understand the needs of all beekeepers.
A comparison of needs between beekeepers in KwaZulu Natal and the Western Cape provinces can be found in Fig. 2. These were addressed as open-ended questions to gather appropriate feedback.
Figure 2: Comparison of needs of beekeepers in KwaZulu Natal and the Western Cape provinces
100 devices were deployed to all 18 participants, with 3G connectivity tested 24 hours before deployment to ensure all devices were active. Data is currently being generated by all devices.
The yellow dots in Figure 3 showcase where the devices were deployed in KwaZulu Natal (Sprint 1), while the blue dots demonstrate where the devices were deployed in the Western Cape (Sprint 2).
Figure 3: In-hive sensor deployment across KwaZulu Natal (KZN) and Western Cape (WCP) provinces
Figure 4: Deployment of in-hive sensors in Western Cape province
AgriSound’s in-hive sensor includes a microphone for collecting sounds from within the hive. Sounds are processed by advanced sound analysis (bioacoustic) algorithms which can be used to provide early warning of key changes within a colony (e.g. swarming events, increased stress, or increased activity). Based on the results from these algorithms, notifications are sent to the beekeepers to take corrective actions.
To determine whether the algorithms would operate as expected under different conditions, we performed spectral analysis of representative recordings of the African honeybee (Apis mellifera scutellate) and Capensis bee (Apis mellifera capensis). Minor differences in the peak frequencies were observed but not enough to warrant full recalibration of the algorithms.
In addition to bioacoustics, the kit also includes temperature, humidity and light sensors which use additional algorithms to provide insights into other key hive metrics including internal temperature (as a proxy for the presence of a healthy queen), humidity (damp / disease risk) and unauthorised hive opening (removal of the hive lid by thieves or predators). These all functioned fully as expected. In addition, sensors were placed in concrete hives (which we hypothesised may have connectivity challenges) and were shown to be able to reliably transmit data to the server.
We also ran a comparison on a giro hive (insulated hive) and a normal hive based on the same pollination site to test gateway communication through insulated materials. The communication was uninterrupted by the insulation, and the Giro hive kept temperatures stable, fluctuating no more than 5 degrees over a 7-day period, measured day and night. In addition, humidity fluctuated no more than 7% over the same period. This contrasts with the uninsulated wooden hive where temperatures fluctuated by a 19-degree range and humidity by a 16% range over the same period. One beekeeper reported a 25% honey yield increase in the Giro hive.
In Sprint 2, events were also attended/delivered in both the Western Cape and KwaZulu Natal provinces. These events aimed to showcase the technology and its potential among relevant industry stakeholders and provide a valuable opportunity to capture feedback from different potential buyers and users of the tech, to inform the development of the business case for the solution in South Africa.
In the Western Cape, an event was held at The British High Commissioner’s Residence in Cape Town on 24th July, with opening remarks from Chris Austin CBE (Foreign, Commonwealth and Development Office) and presentations given by CHAP, CropImpi and AgriSound. Presentations provided an overview of the project, alongside case studies from a number of project participants. A video was also prepared and showcased the technology in-situ, supported by feedback by participant beekeepers. The event was attended by a wide range of stakeholders representing industry, beekeeping, farming, finance, government, industry networks and policy. The event was supported by Aidan Darker and Loice Alusala, both from the Science and Innovation Network, Foreign, Commonwealth and Development Office in South Africa.
In KwaZulu Natal, presentations were delivered by AgriSound at the KZNBFA Growers Symposium (29th July) and KwaZulu Natal Honey Festival, with both events supported by a stand with CHAP and CropImpi.
Figure 5: Attendance at the British High Commissioner’s Residence, Cape Town
Figure 6: Attendance and presentation at the KZNBFA Growers Symposium, KwaZulu Natal
Figure 7: Attendance, presentation and catching up with project beekeeper participants at the KwaZulu Natal Honey Festival
Feedback questionnaires were also distributed to attendees with questions covering usability, affordability, attractiveness and scalability to help inform the business case. Results showed that 80% found the technology to fulfil a need, however, it was notable that only 14% said the technology was affordable. Reviewing affordability and opportunity for different business models will be an essential component of works going forward. Feedback from respondents also noted that they would like to see data in real-time regarding health and disease, productivity, labour saving, business management and theft/vandalism from the devices.
Lessons Learned
During this second sprint, the team was able to validate several critical assumptions that are essential to the overall success of the solution. For example, we have worked with a wide range of different user groups and geographies, demonstrating demand and that the technology works in different climatic conditions. We have also tested connectivity and started gathering representative recording data of the African honeybee (Apis mellifera scutellate) and Capensis bee (Apis mellifera capensis). We have shown that other algorithms (e.g. humidity levels, brood status, theft alerts) are able to function fully as expected, with additional learnings about concrete and giro hives.
Additional questions that will help inform the next sprint include:
How do we make the devices more affordable and what delivery model will be required to achieve this?
Can we gather more data on yield by aligning timing to the nectar flow?
Can the feedback from various stakeholder groups on usability, affordability, attractiveness and scalability help inform the business case and any adaptions required for the South African market?
Next Steps
In Sprints 1 and 2, we focused on training, deployment and initial data gathering from the devices in KwaZulu Natal and Western Cape provinces.
In Sprint 3, we will:
Assess productivity and/or efficiency data from deployed devices and prepare a final questionnaire, which will help inform the overall business case and any adaptions required. We will also compare outputs with regards to demographics, as well as from hardware or software (algorithms), for both KwaZulu Natal and Western Cape provinces.
Explore whether there is a viable opportunity to enter the South African market
If you are interested in learning more about this project, please contact CHAP at enquiries@chap-solutions.co.uk or visit www.chap-solutions.co.uk
About AgriSound:
AgriSound Limited was launched in January 2020 and is based in York. The company was founded by Casey Woodward, an experienced agri-tech innovation expert, with an aim to bring modern technology to insect monitoring. AgriSound has created special smart sensors and listening devices to monitor insects across a range of environments including beekeeping, farming, educational and corporate sites, with the aim of deploying sensors across the planet to transform how people monitor insect activity and make a positive impact on biodiversity.
Since its establishment, AgriSound has partnered with a number of companies and projects, including M&S, Tesco, WWF, Dyson Farming, Innocent Drinks, National Trust Scotland and Royal Horticultural Society to spread the message worldwide and is continually looking for opportunities to branch out further.
For more information, visit the AgriSound website: https://www.agrisound.io/
Instagram: https://www.instagram.com/agrisoundtech/
Twitter: https://twitter.com/AgriSoundTech
LinkedIn: https://www.linkedin.com/company/agrisound/
About Cropimpi:
Cropimpi is an Agricultural Research and Development Company based in South Africa. Cropimpi specialises crop protection trials, demonstration sites for agricultural products, viability projects, business establishment, business development and training. They have assisted many international companies to enter the South African crop protection market by assisting with independent trial work for registration purposes. As part of their social responsibility Cropimpi has an enterprise development section that specialises in teaching rural farmers in their native language to farm sustainably to ensure food security. These charitable training sessions also showcase newer farming technologies like hydroponics for water conservation. Clients include Bayer, Villa, Farmers Agricare, Rolfes Agri and various smaller companies.
About CHAP:
Crop Health and Protection (CHAP), funded by Innovate UK, is one of four UK Agri-Tech Centres. CHAP’s vision is for UK Agri-Tech innovation and expertise to drive sustainable farming systems which deliver economic, environmental, and societal benefits across the globe. CHAP acts as a unique, independent nexus between UK government, researchers and industry, building innovation networks to identify and accelerate cutting-edge solutions to drive incremental, transformative and disruptive changes in sustainable crop productivity.
For more information, contact Senior International Business Development Manager, Dr Jenna Ross OBE — jenna.ross@chap-solutions.co.uk
If you’d like to dig in further…
📚 Explore learnings from the pilot’s first sprint — “Can sensor technologies help to safeguard the future of beekeeping in South Africa?”
⛴️ Read about the pilot’s deployment of AgriSound sensors in the KwaZulu Natal province — “Buzzworthy success: first phase of international project paves the way for future of beekeeping in South Africa”
