Bayer Crop Science: Unlocking value through AI-driven digital farming

How a company is leveraging data science and artificial intelligence for unlocking value in a traditional sector known for stagnating productivity?

In a scenario of stagnating agricultural productivity globally, precision agriculture is a key growth driver for farmers. By optimizing the sowing times and input use, precision agriculture promises a greater return on investments.

Bayer Crop Science AG (BCS), is leading the way in deploying a combination of big data, machine learning, the internet of things, and drones to create and distribute value through integrated digital farming. BCS uses data science to accelerate R&D processes and creates efficiencies in production and supply chain while improving customer experience.

BCS’s Approach to Digital Farming

BCS’s closed-loop approach incentivizes farming automation that reduces time and costs and creates a robust farm-level data pipeline. Big data analytics in conjunction with third-party datasets (e.g., weather) and artificial intelligence then analyzes this data at high speed to funnel back useful insights to farmers for critical, timely, decision-making. In the process, BCS benefits from rich datasets that help it improve its products such as seeds and agricultural biologicals.

Source: The Impact of Digitalization at Bayer Crop Science, 2021.

BCS aims to enhance farmer efficiency by enabling them to remotely and constantly monitor each micro-segment of their plots and determine the precise inputs needed for improving crop productivity. Accordingly, by minimizing overuse, BCS’s digital farming approach also furthers the cause of sustainable agriculture[1].

Building a data pipeline

BCS collects data from farmers’ fields through their equipment, in-field sensors, and drones, and triangulates it with third-party data from academicians, governments, and its employees. For the farm-specific data, BCS’s smart drones deploy special imaging technology to identify granular crop growth patterns and point out crop stress in terms of diseases, drought, or pests much before it is visible to the naked eye; which helps farmers take preventive than curative measures. BCS has also partnered with the world’s largest drone service provider, XAG, for digital farming and precision spraying in Japan, Southeast Asia, and Pakistan[2]. Secondly, BCS’s in-field sensors enable farmers to monitor soil health in terms of moisture, nutrients, salinity, etc. in real-time and together with geo-location data, enable them to optimize input use. Lastly, through smart combines having sensors and video imaging features, BCS enables farmers to collect rich data. For instance, precision planters couple location data with past and forecasted field trends to plant seeds at the right location, spacing, and depth[3].

Delivering insights-based value to farmers

BCS’s digital applications like Climate FieldView™ leverage AI and billions of data points collected through its robust data pipeline to help farmers know which right product to apply, where and when, at in what quantity specific to their fields. Apart from maximizing yields and thus revenues, it also helps farmers cut costs, save time, and conserve inputs, including soil.  Similarly, through image recognition, Climate FieldView™ enables early detection of diseases before they are manifested as wilting, browning, molding, or rotting and suggests mitigation measures.

Source: Introducing Climate FieldView at Bayer Crop Science, 2021[4]

“For precision planting and precision farming, companies like Bayer work really closely with the farmers to understand their land, their acreage, the type of soil they have, the water flow, and then suggest the best seeds and best plant breeds that will thrive in those conditions” says Michelle Lacy, data strategy lead for R&D in the Plant Biotechnology Division at Bayer Crop Science[5].

Capturing value through AI-driven product R&D

BCS’s closed-loop process generates rich data that it leverages for accelerating the pace and success of plant breeding innovations by accurately predicting genetic outcomes during trials. Scientists use machine learning to simulate a new plant variety’s performance in thousands of micro-level agroclimatic and soil conditions, helping breeders create a thoroughly vetted product at a much lower cost. Additionally, Nalini Polavarapu, Enterprise Data Science Strategy Lead at BCS says “We were able to save an entire year of testing in our pipeline by using machine learning”[6].

Secondly, in partnership with Google Cloud, BCS leverages ML to breed seeds that are custom-designed to thrive in specific regions, climates, and soil types. Plant height, color/greenness, and plant diseases/infections as captured by images are triangulated with soil, irrigation, and fertilization data to identify the most resilient breeds[7].

Through an AI assistant created for breeders, BCS allows them to identify condition-specific traits of resilient varieties, and then suggest where exactly in the plant genome can they find the associated genes. Additionally, post disasters or diseases, BCS’s drone images help estimate crop damage to aid in insurance claims processing[8].

Source: Data Science in Plant Breeding at Bayer Crop Science[9]

Lastly, apart from precision product design, BCS characterizes the soil properties of their internal testing network and conducts environmental similarity calculations and historical modeling to establish a representative global testing network.

Responsible Innovation

To sustainably drive innovation, BCS claims strict adherence to FAIR (findable, accessible, interoperable, and reusable) data principles; allowing users to be in control of their data. “It’s extremely important,” Lacy says. “It’s the foundation of our data strategy”[10].

In summary, Bayer Crop Science’s holistic approach to seamless on-farm data creation and collection, machine learning deployment to generate valuable insights for real-time precision farming decisions, and leveraging data for R&D are transforming the productivity frontiers in Agricultural ecosystems.

[1] “Digital Farming Systems,” accessed October 4, 2022,

[2] “Bayer and XAG Collaborate to Bring Digital Farming Technology to Smallholder Farmers in Southeast Asia & Pakistan,” accessed October 4, 2022,

[3] “Digital Farming,” accessed October 4, 2022,

[4] Introducing Climate FieldView, 2021,

[5] Senior Writer, “Transforming Analytics into Business Impact,” CIO, accessed October 4, 2022,

[6] “Machine Learning in Agriculture,” accessed October 4, 2022,

[7] “Improving Soil Health and Crop Management with Geospatial Analytics on BigQuery and Dataflow.,” Google Cloud Blog, accessed October 2, 2022,

[8] “Data Science in Plant Breeding,” accessed October 4, 2022,

[9] “Data Science in Plant Breeding,” accessed October 4, 2022,

[10] Writer, Senior. “Transforming Analytics into Business Impact.” CIO. Accessed October 4, 2022.


Marriott: Data-driven Customer Experience for Decades at Scale


Leveraging Big Data at Ibotta

Student comments on Bayer Crop Science: Unlocking value through AI-driven digital farming

  1. Super interesting casse Anand – thanks a lot!

    The idea of bury real time sensor into fields makes perfect sense and seems highly scalable too me. As far as I understood it out of the second video, Bayer Croop Science is mainly operating in Britain and Western Europe but has now partnered with the world’s largest drone service provider, XAG, for digital farming and precision spraying in Japan, Southeast Asia, and Pakistan. This got me thinking. I agree with Bayer that the idea of digital farming would generate more added value in less developed countries such as Southeast Asia or Africa. However, I wonder if the use of drones in these regions is scalable. If you think about it, every field on the planet would be a potential target. How do you successfully get such delicate objects, as drones are, into a rough field and from there across the country without breaking them. I also wonder if every farmer in this area has access to a tablet or smart phone. Has anyone information about how Bayer plans to handle this or are they currently just focusing on large scale industry fields?

  2. Thanks Anand, this is fascinating and clearly shows how technology and data could create value in developing countries. I had almost the exact same questions as Yannik’s while reading through the post. I wonder if certain economy-wide factors (lack of quality input, low adoption of technology, climate change etc.) are restricting Bayer’s ability to scale. But overall, I feel very encouraged by this post knowing that this type of companies are creating values around the world!

  3. Thanks very much for this fascinating post, Anand!

    I’m most curious about the data labeling process for farm-specific data. It strikes me that being able to use special imaging technology to identify granular crop growth patterns and point out crop stress in terms of diseases, drought, or pests is in and of itself a remarkable use of ML. I wonder how long it took to train that component of the model and how many drone flights it required…

  4. Loved the article! I was surprised to see this initiative from Bayer as I have heard about the technology times before, more from startups who are entering the industry and disrupting the way farming is done, in a more climate friendly and sustainable way. And this is somehow hurting Bayer’s business who is also selling insecticides and by using AI and drone assisted technology for farming, the use of such chemicals would be reduced. Do you have any insights in their thinking and approach? Also, was this technology developed inside, or are they acquiring innovative start-ups? And how are they planning to compete against them going forward?

  5. Thanks for this lovely post Anand! It was a joy to read and it was particularly fascinating to learn about application of AI in such a niche, but socially-impactful domain!

    Apart from the obvious value created to the farming sector by use of ML, I wonder whether Bayer also gains by use of data to train models, uncover insights that it can apply to completely different application domains.

    — Aditya Mate

Leave a comment