Oracle AI & Data Science Blog
Learn AI, ML, and data science best practices

How Oracle and The World Bee Project are Using AI to Save Bees

Justin Charness
Director, Product Marketing- Oracle AI

Oracle recently announced a partnership with the World Bee Project to leverage cloud technologies to better understand the decline in bee populations globally and devise innovative strategies to help farmers manage bee and pollinator habitats. I sat down with Andy Clark, Design Innovation Director for the Cloud Platform Innovation team to tell me more about the project.

Justin: Tell me about the Cloud Platform Innovation Team at Oracle and what kinds of projects you focus on.

Andy: The Cloud Platform Innovation Team at Oracle is a newly formed group, based in the UK, which works with customers across industries who are trying to innovate with cloud technologies but are struggling to execute their ideas in reality. We leverage a “design thinking” approach to unearth specific business needs that our customers are trying to meet and then work closely with our customers to frame the business challenge and ideate potential solutions. We then build specific prototypes which the customer tests in their market to prove the business value.  In parallel to our work with clients, we work on a number of broader innovation projects that showcase the power of Oracle technologies, including the Bloodhound Project, and now the World Bee Project.

Justin: What is the World Bee Project? What are they trying to accomplish?  

Andy: We’ve all heard the news that there is a problem with the decline in bees, but most people don’t really know what that means or how severe the impact is. Around 77% of the world’s food supply depends on pollinators (bees, butterflies, etc.) In addition, pollinators play a primary role in the reproductive cycle of plants and trees by spreading pollen and seeds. It’s not just about the food we eat, but the air we breathe and our entire ecosystem.

The mission of the World Bee Project is to increase food security and livelihoods by combining AI and sensor systems with world-leading bee research to provide farmers and the general public with the knowledge and solutions they need to foster healthy habitats for pollinators. The project is led by Sabiha Malik, who has brought together government agencies, leading researchers, and the private sector to attack this problem from all fronts. Our team started working with Sabiha last summer to map out the right technology stack to help attack this problem using the latest emerging cloud technologies.

Justin: How did your team get involved? What approach are you taking to solving this problem?

Andy: We approached the World Bee Project using the same process as our innovation projects with Oracle clients. The first step was clearly identifying their higher-level goals and then mapping those goals to specific deliverables.

The first major goal was finding new methods to collect and analyze data from the bees and hives themselves. Most other Hive research was happening locally, so we knew we needed to find new ways to collect data and then manage and analyze that data across hives on a global scale.  What this came to was working with a number of specialist sensor providers and putting their sensors on Beehives then connecting them to the Oracle Cloud. Through these sensors, we’re able to capture a wide variety of data points including temperature, humidity, pollinator decline/deficits, and acoustic data (ie. bee sounds). We’re capturing this data from hives across global regions and then integrating it with freely available data around weather, mapping, and other relevant local data. We’re then using artificial intelligence, analytics, and data science to look for patterns and anomalies in bee colony health in certain regions in the world to identify problem areas.

Justin: Can you tell me more about the Oracle Cloud tech that’s involved in the project?

Andy: Other than the sensors, the whole tech stack for the project is in the Oracle Cloud. We needed to design a system that could handle huge volumes of data coming in in real time across global hives and which we could process efficiently in the cloud. We’re reading all our data into the Oracle Autonomous Database and we’re leveraging the Oracle Analytics Cloud for machine learning modeling, and also for data visualizations. We’re also leveraging Oracle Cloud Infrastructure for compute and data storage and processing. It’s a tightly integrated cloud system that allows multiple global teams, including leading research institutes such as the University of Reading, to analyze the data and understand insights efficiently.

Justin: What does the future of the project look like? What are some concrete solutions you’re exploring?

Andy: We’re beginning to get enough data from the hives where we’re starting to see anomalies and uncover new insights across global regions. We’re also exploring ways to present the data to farmers in a way that’s understandable and actionable through dashboards or other real-time reporting tools so they can improve the everyday management of their colonies and improve productivity and honey yields while combatting the spread of disease, parasites, and pests.

One particularly interesting area is the analysis of acoustic data that is gathered from the hives.  You can tell an enormous amount about the behavior and health of a bee colony as well as its relationship with the surrounding environment from the “hum” that it makes.  For example, we can predict from the hum when a hive is likely to swarm and warn a beekeeper or farmer.  A swarm is a perfectly natural occurrence that happens when the bees have outgrown their hive and involves roughly two-thirds of the bees leaving with their queen to start a new colony leaving a new queen to continue the existing one.  It can be a problem, though, if the bees swarm at the wrong time of the year, for example in the winter, because they are under threat from a predator or disease.  The beekeeper will also probably wish to capture a swarm and steer it to a new hive thus increasing the number of bees and honey yield they will have.

We’re also exploring a number of innovative solutions using emerging technologies that address the bee/honey productization lifecycle to certify and promote sustainable farmers and farming practices that promote healthy pollinator habitats. The goal here is to start making it financially attractive to farmers to adopt sustainable practices, and also to satisfy consumers who are concerned with the decline of bee populations and care about the provenance of the honey and other pollinated produce they buy.

One area we’re looking at is the use of blockchain across the supply chain using the honey signature itself.  Honey has a very unique signature which is based on the types of pollen and nectar the bees collect, which is based entirely on the habitat that surrounds the hive.  You can take a sample of honey and break down the DNA of that sample and store it in the blockchain. What this allows you to do is certify that the honey hasn’t been tampered with across the supply chain and its journey to the retailer. This enables you to provide consumers with a product that is truly sustainable and is promoting healthy bee habitats.

Another area we’re looking at for an eco-label is certifying farmers on the basis of the farm makeup itself. Farms that plant a percentage of their land with flowering crops such as spices, oil seeds, medicinal, and forage plants see significantly increased crop yields by as much as 79% as a result of more efficient pollination from bees.  What we would do here is take aerial images of the farm itself with drones or even satellite images and then use image recognition algorithms (AI) to evaluate whether farmland is allocated in a way that ensures the prosperity of their bee colonies and other pollinators.

We’re also investigating what we can learn by monitoring the actual movement of the bees both in the hive and across the surrounding environment by analyzing video footage.

These are still areas of exploration, but there’s enormous potential to leverage emerging technologies and the cloud to drive the entire business ecosystem towards one which promotes both bee and farmer prosperity.

Justin: If people want to learn more about the project or about the Cloud Platform Innovation team at Oracle, who should they contact?

Andy: for questions on the innovation team, you can reach me directly andrew.w.clark@oracle.com. For more information on the project, visit worldbeeproject.org. If you are at Oracle Open World in London, please stop by our booth where we will have a live demo of our solutions using Oracle Cloud technologies.

For more information on Oracle's AI products and services, please visit http://www.oracle.com/ai




Be the first to comment

Comments ( 0 )
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.