The next wave of advancement in agriculture will have elements of big data, says precision ag expert
By Jackie Clark
Staff Writer
Farms.com
To take full advantage of improving technological tools in agriculture, producers will need to be able to farm their land, but also farm their data. This was a key message Dr. Raj Khosla delivered to 2020 Farms.com Virtual Precision Agriculture Conference & Ag Technology Showcase attendees on Nov.17. He’s a professor of precision agriculture at Colorado State University, and founder and past-president of the International Society of Precision Agriculture.
Much has changed since 1983, when “global positioning systems (GPS) became available for public use,” Khosla explained. In 2000, GPS accuracy improved from 100 metres (328 feet) to 15 m (49 ft.), and “that was a game changer because it allowed agriculture to become very innovative.”
When location-based agriculture started to advance more rapidly, “founders of digital agriculture cautioned us. They said once we start engaging ourselves in location-based agriculture, we will embark upon a new journey. And that new journey is collection of data,” he said.
Now, farmers can collect many different layers of data from a single field, every year, capturing variability in measured parameters over space and time. We have the technology to gather that information, but how do we turn information into effective decision-making, Khosla asked.
In other words, “we know how to farm the land, but what we don’t know is how to farm with data,” he said.
Over the years, farmers in North America have made “dramatic progress in our ability to produce more from the same unit of land,” Khosla explained. Digital agriculture can be an integral part in continuing that trend.
“The data that we are collecting is only as good as how we know how to translate that data into information using robust algorithms into an improved decision-making process,” he explained.
The first wave of digital agriculture involved GPS, yield monitors, and some variable rate technology, whereas, the second wave added auto guidance, proximal and remote sensing, robotics, and refined variable rate technology, Khosla explained.
“What does the next or current wave of digital agriculture look like?” he asked. “It has elements of big data. Now, data alone doesn’t tell you anything unless you perform robust algorithms or analytics.”
If producers “farm the data, it enables us to add value,” he added. “We all know that when we add value … it fosters adoption.”
To improve management to add value, farmers must be able to easily collect data and translate it into management decisions.
“You can only manage what you can measure. So, it becomes very important in data-centric agriculture, that we must have the capability to measure,” Khosla said. However, agricultural systems are very complex, with a multitude of measurable factors that influence crop growth and development.
Using the example of one field with a centre pivot irrigation system, a farmer would have the ability to adjust water at the nozzle level, meaning they could have over 5,000 water management zones in that particular field.
In that case, “I have the technology. What I don’t have is the science to translate into a prescription that will allow me to make the best management decisions, in terms of water, in this case,” Khosla explained. So, “for us to take advantage of digital agriculture, we have to embrace a whole network of sensors and sensor data.”
Khosla’s lab has been developing the next generation of small sensors that can be arranged in a network to measure soil moisture and nitrates across a farm field.
“Your entire field is going to look like the Internet of Things,” he said. And, the sensors are biodegradable within a year “so you don’t have to worry about going back to the field and collecting the sensors.”
The project is called Precision Agriculture using Networks of Degradable Analytical Sensors (PANDAS), and Khosla’s team hope the sensors will become commercially available at less than a dollar per sensor.
The project is promising, however, the complexity of agricultural systems means that there are any number of additional factors that sensors could measure.
“We’re trying to understand what data layers are important and pertinent,” Khosla said. “We’re not there yet – there are many things that we don’t know.”
Researchers who are working on this problem are investigating the relative importance of each layer of data they’re measuring, how they each contribute to outcomes, and how to account for seasonal anomalies.
“We need to make sure it’s robust, it’s reproducible, and it’s reliable,” Khosla said. Once the data collection and management algorithms are all those things, precision agriculture can involve automated decision-making, with the ability for the farmer to override if they wish.
A few other challenges exist, such as finding expertise in digital agriculture, big data, and agricultural systems, Khosla explained. Also “rural broadband continues to be a roadblock when we talk about translating digital agriculture into a reality.”
However, Khosla is optimistic for the future ability to farm data to improve agricultural production with reduced inputs, and to continue to grow more with less.
William_Potter\iStock\Getty Images Plus photo