Meet the AgriWebb R&D group part 2: The need for AI in AgTech

Hi all, Kenny Sabir, AgriWebb's VP of Research & Development, here again to dive into Part 2 of our Think Series, Meet the AgriWebb R&D Group . As you may have already guessed from my AI-generated portrait above (look at those robots! It's almost like my AI tool knew what I wanted to talk about this week!), I'm diving into a topic that's seemingly omnipresent these days: Artificial Intelligence.

You've probably had a hard time escaping artificial intelligence (AI) in the news recently, whether that be the sudden popularity of ChatGPT for school kids doing assignments with ease to lawyers referring to bogus prior cases that the AI made up . Or maybe you heard about graphic designers and artists at danger of losing their jobs due to the derivative art coming from AI projects like Stable Diffusion and MidJourney that can transform a simple text description to a highly detailed artwork. While certain jobs have been headed toward obsolete while new fields have in turn emerged over the past 400 years, the speed at which AI is affecting the existing workplace is staggering, with no slowdown in sight. However scary this may feel, there are many cases where AI is helping accelerate business development and intelligence, too. So the question arises: when it comes to livestock farming, can AI be used for good? Below, we'll dig in to one of the biggest questions on many of our minds today.

Remember: you're already collecting the data

Livestock agriculture plays a crucial role in meeting the world's demand for protein. However, as the global population continues to grow, so does the need for sustainable and efficient livestock production. We need to develop new ways to become more efficient, while making farms more resilient to changing weather conditions. As the most effective producers know, the first step of improving your business is measuring it. Farmers and ranchers do love their data; they've been collecting it for years, whether that be in the notebook in the top pocket, or meticulous rainfall records in logbooks that have been handed down for generations. 

With the advent of AgTech for livestock production , this important data, including paddock movements, inventory and treatment records, feed and weight records, birth and death, sale and purchase, tag and score, can be entered on your phone while you're in the paddocks. When you're back in range, it'll automatically synced back to all workers on your farm who have a login to the same livestock management software platform you do! When you have this type of digital track record, it's not only easier to communicate – your data also isn't at risk of getting lost when the notebook goes through the wash. Plus, apart from helping you to quickly analyse and understanding the data behind different scenarios on your farm, livestock management software solutions are also a perfect record capture for succession planning and audit preparedness. 

Automation is now providing so much more than data entry alone

Producers have long known that government weather stations often miss the actual weather on farm, hence the recent boom in IOT devices such as personal weather stations recording localised rainfall. Satellite data , known as remote sensing, is available to help producers know how much dry matter is available in the paddocks for feed that is updated every 5 days. Producers have been keen to sign up to remote sensing over the last few years and according to the latest State of the Global Farmer survey , bringing in remote sensing data is the highest sustainability priority. There are more benefits to remote sensing: now groups like Cibo Labs can also automatically determine your ground cover and tree cover (apart from the Feed On Offer they are well known and loved for), while FlintPro can understand your above ground vegetation carbon, biodiversity and water analysis without needing an on-site visit. Now that we have access to all this varied and detailed data, how can we understand trends to make better decisions on-farm?

So many inputs!

Researchers have been working out biophysical models in livestock and pasture research for decades, coming up with the particular equations to handle building growth curves based on a set of inputs. But how can we get a solution that works for the livestock farmer population of an entire country, given that many academics have typically done their thorough testing in a limited geographical area? The reality is, pasture modeling is very complex when you consider the variety of regional rainfall, additional climate data, pasture types, soils types, and topology. Likewise, livestock growth can be affected by the feed availability, weather, diet changes, distance moved, sickness and more. All of this is highly reliant on the future weather, for which there is no crystal ball, though there are probabilistic forecasts and decades of historical data to analyse. 

The good news is we don’t have to come up with a really complex equation that represents all these inputs; we can now feed them all into ‘machine learning,’ which can detect patterns and create prediction models for us. Nature provides inherently messy data sets, which machine learning can take into account to give probabilistic results so you can make smart on-farm decisions that take all the information available into account. These models can be highly localised, to both your region and also to unique historical data on your actual farm, making good use of all the data you have been collecting (hopefully, with AgriWebb !). The best bit about ‘machine learning’ is in the name; the longer it runs, the smarter it gets. As new weather, pasture, and livestock data become available, machine learning will fine-tune itself to understand how to better predict for a given location. 

... and now comes AI

Machine learning is just the first step toward the AI landscape that is fast evolving all around us. From machine learning, we can move on to AI to analyse a "digital twin" of your farm, that can help you understand the impact of different management decisions in a scenario planner by enabling you to run simulations and optimisations to your heart's desire.

"What is the financial risk if I buy animals in 2 months and the weather turns?"

"How can I plan stock rotations given good and poor weather scenarios over the next 4 weeks?"

"What is the cost benefit analysis given different weather scenarios between buying feed during a dry period to reach a target weight, vs cutting my losses and selling early?"

"What is the impact that each of these decisions have on my farm’s natural capital?"

AI isn’t taking over the farm, it's just helping you farm better. You can use the right AI as a tool like any other to help provide data-driven insight so you can make better business decisions.

Livestock shouldn't be left behind this time

For many years, AI has been enshrined in the domain of computer scientists and science fiction, not democratised as relevant or accessible to most people. However, with the rapid growth we are witnessing when it comes to consumer AI products, it will soon be a part of the common vocabulary. People won't just want it, they'll expect it! In just the recent past, the livestock agriculture industry experienced a parallel problem with technology. After the invention of a new generation of Smartphones in 2007, it took another 7 years before producers had a decent phone application available to them to record their farm data. During this period, many other industries had already modernised due to the wide range of producers and solutions already being available to them to leverage and learn with. Over this period, producers also took to iPhones and Facebook lamenting how the livestock industry was slow to adopt useful technology while their cropping neighbours had already been taking advantage of GPS tractors for 20 years. We shouldn't make the same mistake this time around.

As AI booms, AgriWebb and many of the amazing researchers and AgTech innovators we work with are all ready embrace AI and find ways we can leverage it for the benefit of producers everywhere.

What do YOU think?

That may be it for this week's post, but I'd love to hear your own take on AI and livestock agriculture in the comments below! Drop a line, and let's keep the conversation going And if you have your own amazing AI-generated portrait to share, you can always add that, as well! ;)