Next-Generation Field Trials: Leveraging Agricultural Analytics

“Big data” has become a behemoth in tech and in everyday life. It’s a term so common that a quick Google search returns 7,3000,000,000 results. Big data refers to extremely large datasets that require next-level architecture to manage.

Controlling agricultural robot – illustrative photo. Image credit: ThisisEngineering RAEng via Unsplash, free license

Big data is a fact of life in many ways, influencing how we connect with others on social media, how music and videos are tailored to us on the Internet, and even how police are stationed to keep us safe. Big data has future applications, too – providing information to machine learning and artificial intelligence. 

There’s one example of the use of big data that might not be top-of-mind, but touches the day-to-day lives of every person on the planet: agriculture! Large, complex datasets are used in the agriculture industry to improve decision-making throughout the growing season. Big data is used to make key, yield-influencing decisions like when and how much fertilizer to apply, which varieties of a crop to plant, and even when the ideal time is to harvest a crop.

The power of big data in agriculture will only become more impactful as the need for more food to be produced on less land is driven by an increasing population. The world population is projected to reach nearly 10 billion people by 2050, meaning there will be more mouths to feed on the same – or less – arable land. Global food security hangs in the balance.

To address this critical challenge, the FAO estimates that overall food production will need to increase by 70%. Achieving this without any additional land and limited natural resources might feel impossible. But big data analytics in agriculture is one way to help increase yields and profitability for farmers while also protecting the environment – and securing the future.

From decisions at the field level to decisions at the farm business level, insights from big data analytics have big potential to make a difference.

Harvester working in a farm field – illustrative photo. Image credit: Luke Thornton via Unsplash, free license

Big data analytics in agriculture

Farmers have long been using data to make decisions. But only recently – in the last 200 years – has the scale enabled farmers to move beyond in-season decisions to multi-season and whole-farm decisions. Tools, resources, and expertise in agriculture have made the use of data on-farm even more efficient.

Off the farm, the agriculture industry has used data to drive innovation. The industry has also benefited from the exponential increase in data scale. Across the agriculture industry, big data applications can take shape in many ways – but nearly every aspect of agriculture is touched by technology and data science today.

Creating insights from big data requires large datasets, systems powerful enough to process them, and the capacity to extract those valuable insights. Big data can be distinguished by the 4 V’s: volume, velocity, variety, and veracity.

• Volume: There is an enormous volume of data available in agriculture, and that continues to grow. Farmers are collecting data all the time; this can happen manually, through applications on mobile devices. Or, data can be collected automatically through machines equipped with sensors that communicate with servers on the Internet. This creates the Internet of Things. Imagine the size of a dataset that includes sub-acre information collected from multiple passes over the same land each year – for decades!
• Veracity: Data quality is critical to enable meaningful insights from large datasets. Without high-accuracy data, the trustworthiness and value for scientific meaning are minimal. Veracity includes the processes used to clean data of irregularities.
• Velocity: Just like it sounds, velocity is about speed. Specifically, the speed at which data can be collected, sorted, processed, and stored – which has now near-real time. Velocity has only increased over the years, though data can be collected at different speeds, and lends itself to be very useful for in-the-field evaluations. For farmers, this means getting critical feedback while there’s time to make critical decisions, or getting a real-time scorecard of crop performance as they harvest a field.
• Variety: There are many different types of data assets – and the different kinds keep growing. Types of data can include data from sensors, structured databases, video streams, and logs, for example. In agriculture, it’s common for data to be collected in multiple ways on the same land. For example, a grower might use drones for big-picture, aerial views of crop emergence or disease. But for details, sensors in the soil or on machinery can show fertilizer needs or soil saturation. And data inputs might be logged manually for weather or economic trends.

Some consider other factors in the definition of big data. Volatility, or the pace at which data changes or has a usable life, can determine the storage needs of the data. Visualization refers to conveying insights from data patterns through images, graphs, or infographics. And value is really about the objectives of the analysis. In agriculture, the value is derived from short-term insights that support decision-making that aligns with long-term objectives like sustainability or profitability.

Big datasets can be combined from multiple data streams or created independently. Standardizing the datasets ensures that each data point is formatted similarly, is consistent with what’s being measured, and has proper labels.

This standardization is key to enabling analysis. Combined with the 4 V’s mentioned above, standardization is what gives big data value. Without it, data is siloed and limited. Focusing on efficient and streamlined processes for dataset generation and management through standard operating procedures is a good investment now and in the future. Best practices put in place ahead of data generation can save tons of time in standardization down the road.

The vast potential of big data in agriculture

Depending on the area of interest, industry professionals can use big data for various evaluations. Whether it’s agronomy or economics, natural resources, or food science, big data can support a deeper understanding of the industry and how to improve productivity.

Big data analytics allows people across the industry to identify meaningful results that deliver informed agricultural decisions. The higher quality of the data, the more powerful the results – because confidence in the probability of a true effect increases.

Collaboration across the industry takes this opportunity even further. Standardized data can be shared once collected, and multiple people can perform analysis and evaluation – even for different purposes. Big data in agriculture is supporting innovation and decision-making at input companies and food and beverage companies and for farmers, and researchers. 

• Agriculture Input Companies: Big data and agronomic field trials are used to evaluate the efficacy, safety, and marketability of new products. For companies that produce seed, fertilizer, crop protection, and machinery, agronomic field trials and the data they generate support innovation and commercialization.
• Food and Beverage Companies: For these companies, quality and yield are top of mind. Through crop nutrient planning and working with growers and big data, they want to ensure their raw ingredients – crops – are produced sustainably while delivering the expected quality.
• On-farm Experiments: Farmers use agriculture big data analytics to maximize productivity and increase production sustainability. This ensures their futures.
• Researchers in Universities, NGOs, and Government Research Centers: Agronomic field trials are a tool researchers use to test hypotheses and experiment with solution options. This work addresses challenges farmers face around the world and close to home.

The research questions and experiments for these stakeholder groups focus on a variety of topics. One might be seeking information on disease-resistant crop varieties and another might be looking to understand the impact of nutrient placement on crop yield. Answering these research questions requires large amounts of data for statistical analyses.

But there’s one group in particular who get to directly connect data to decision making – and that’s farmers doing on-farm experiments. Let’s dive deeper into on-farm experiments.

Big data and on-farm experiments

Big data plays a role in many areas of agriculture. One place where data directly connects with decision-making for the future is in agronomic on-farm field trials. This experimentation involves testing out practices, products, and equipment in real-world scenarios to understand their effectiveness in that cropping system. In order to effectively evaluate an agronomic field trial, data must be collected.

In the process of on-farm experimentation, farmers may gather data on crop growth, weather, topography, disease and pest pressure, soil moisture, and more. This data can be pulled into statistical models or machine learning. Big data analytics help inform best practice decisions for the future, to improve yield or minimize risk. Data analytics provides the ultimate report card for the experimentation on the product, practice, or equipment.

Collecting, managing, and analyzing data from agronomic field trials isn’t easy for researchers or professionals; it’s definitely not simple for farmers, who are trying to run a business and day-to-day farm operations. But the value of on-farm experimentation ties back to the data collected. This is where trial management systems can help.

At the farm level, software that connects farmers and agronomists with data management and analysis tools can be game-changing. Trial management platforms can help standardize data, complete statistical analysis, and even collaborate with others. These tools make big data analytics accessible to anyone in agriculture.

Collectively, on-farm experiments can be enhanced through the use of big data alongside next-generation technologies and solutions. Next-generation on-farm experiments and field trials will retain the most effective aspects of historical field trial processes, but leverage data and cutting-edge technologies to improve decision-making and ultimately enhance productivity.

On-farm experiments can become even more powerful, collaborative, and effective through cloud-based, next-generation solutions that harness a data approach and AI-driven recommendations. These trials will go further than ever before to ensure a sustainable future.

With the need for bigger yields in coming years, it’s essential that farmers harness big data analytics to increase productivity year after year. It’s a challenge to do this while farming sustainably and staying profitable. Big data analytics help translate agronomic field trial data into useful insights that position farmers, agriculture, and the world for a healthy future.

Author Bio

Ron Baruchi is the president and CEO of Agmatix. With over 20 years of experience in the technology sphere, Ron has taken this experience to the agricultural sector. Passionate about using data to solve complex problems, he has used his expertise in technology with Agmatix to improve crop yields and quality while limiting environmental impact.