How artificial intelligence is paving the way to a sustainable farming future

When in 1955, John McCarthy, one of the founding fathers of artificial intelligence, described it as “the science and engineering of making intelligent machines”, he could not have envisaged its use today in a myriad of applications from Amazon’s Alexa smart speakers to facial recognition technology used in smartphones.

Almost seven decades later, in his GatesNotes blog, Microsoft co-founder Bill Gates hailed the beginning of the Age of AI. Artificial intelligence will “change the way people work, learn, travel, get health care and communicate with each other. Entire industries will reorientate around it, and businesses will distinguish themselves by how well they use it,” he said.

But what is AI? Put simply, it is a branch of computer science that aims to mimic human intelligence. It is about building machines able to think and act intelligently and work by combining large amounts of data with superfast processing.

Currently agriculture is one of the least digitalised industrial sectors, but development and use of AI is rapidly gaining pace, says Bayer UK digital manager Max Dafforn.

AI offers the potential to boost productivity, cut waste, reduce costs and help tackle some of the major challenges farmers are facing across the globe – climate change, a diminishing workforce, plateauing yields – against a backdrop of a growing world population and the need to produce food more sustainably.

Bayer’s focus is on harnessing the power of AI to help support farmers and growers in the transition to more regenerative farming systems, investing in innovations to increase food production, farm incomes and climate resilience, while also protecting and renewing nature.

“A regenerative agricultural system that produces more with less is only possible with advanced digital technologies that use artificial intelligence and data science in every phase of the farming cycle,” Frank Terhorst, Head of Strategy and Sustainability in Bayer’s Crop Science division, told visitors to farm machinery show Agritechnica in November 2023.

“Agriculture today is infused with technology, whether in seed breeding and crop protection discovery, during planning, planting and harvesting on the farm, or in virtually all interactions with value chain partners,” he said.

Plant breeding

In Bayer’s global plant breeding business, artificial intelligence in the form of predictive analytics and machine learning is being used to select new breeding crosses and predict breeding outcomes on a scale and at a speed that has never before been possible.

“As we were scaling our breeding and field research efforts and starting to collect more dimensions of data, it became virtually impossible for a human being to sift through the data to make informed decisions,” says Phani Chavali, who leads Bayer’s plant breeding analytics team.

In response to this challenge, Bayer built an artificial intelligence assistant that helps breeders select the right candidates in the breeding program. It relies on cloud-based algorithms built on a foundation of roughly 1.7 trillion calculations, enabling a dramatic shift in the scale and speed of the breeding pipeline.

On-farm solutions

Meanwhile, on-farm, machine learning, AI, and data science are delivering new levels of precision through Bayer’s FieldView digital farming platform and range of innovative digital products.

MagicTrap, a digital yellow trap for oilseed rape, is due for launch in the UK this summer. A yellow water trap with a difference, MagicTrap features a high-resolution camera which regularly photographs the trap’s contents, automatically identifies and counts the insects present, and alerts growers via a smartphone app. At a stroke, this eliminates the need to physically inspect traps and provides continuously updated intelligence on pest pressures, helping to inform integrated pest management decisions. The system can currently identify cabbage stem flea beetle, pollen beetle and weevil, and will expand to other pests in the future.

With lack of data connectivity identified as one of agriculture’s biggest challenges, a strategic collaboration between Bayer and tech giant Microsoft has seen the launch of Azure Data Manager for Agriculture and cloud based AgPowered Services, designed to support data interoperability, improve transparency and accelerate innovation.

One area currently being developed is the use of expert generative AI systems to benefit farmers and agronomists in their daily work. Bayer is piloting a large language model (LLM) trained on years of proprietary data which will harnesses GenAI to quickly and accurately answer questions relating to agronomy, farm management and Bayer agricultural products.

A separate GenAI prototype has been developed which will allow users to interact with and receive actionable insights from their farm data via a chat interface. “

This will allow farmers to query their own data using natural language. So instead of having to trawl through spreadsheets or lots of records, you can simply ask it a question, for example: Which of my fields had most nitrogen last year and how did they perform compared to others?,” says Max.

Crop protection

Bayer is also using artificial intelligence to transform its approach to developing crop protection chemistry, applying AI to move from incremental improvements in existing chemistry to breakthrough innovations based on designing new chemistry with specific performance and safety profiles.

“We call this the CropKey approach,” says Max.

Crop protection works by finding and inhibiting a given protein in the target organism. CropKey uses AI and machine learning to identify these unique protein ‘locks’ faster and with greater accuracy than ever before.

The active compounds that are the ‘keys’ to these locks were until recently physically screened from Bayer’s 2.6 million compound library and selected according to their potential efficacy as crop protection solutions. Today entirely new and precise keys are being designed using virtual screening and computational predictive modelling.

These data-driven technologies narrow down those compounds with promise of efficacy and eliminate those that do not reach pre-defined safety and sustainability standards far more quickly and accurately than previously imaginable. When a perfectly designed key fits a unique lock, Bayer’s scientists have found a new mode of action.

CropKey is already producing tangible results, with more than 30 potential new molecular targets under investigation and more than 10 targets validated as new modes of action in early research. Two entirely new modes of action have been designed using the approach: a unique broad-spectrum fungicide for use in fruit and vegetables and a new herbicide molecule for broadacre post-emergence weed control.

As UK agriculture transitions away from area-based payments, the AI tools Bayer is developing will help farmers and growers get the best possible value out of each and every input they are applying.

“There is no doubt in my mind that AI is going to have a huge impact on agriculture. I would encourage farmers and agronomists to engage, explore the technologies and look for opportunities that are going to best work for them and benefit their businesses,” says Max.