Precision Agriculture Can Enable Continued Intensification and Boost Yields

 

     According to Wikipedia precision agriculture “is a farming management strategy based on observing, measuring and responding to temporal and spatial variability to improve agricultural production sustainability.” Amanda Ashworth of the USDA’s Agricultural Research Service described precision agriculture as follows:

 

“Precision agriculture is a general term to describe farming tools based on observing, measuring, and responding to within-field variability via crop management. It is made possible through the use of Global Positioning System (or GPS satellites) or Global Navigation Satellite System (GNSS), which enable farm managers to respond to field irregularities. This approach allows farmers to make important resource management decisions both on-site and in real-time.’”

 

     Automation, robotics, and sensors are utilized in precision agriculture. Drones, GPS, and GNSS (global navigation satellite system), satellites, and remote sensing technology may also be utilized. Precision or automated agriculture can improve and optimize many agricultural processes. This can lead to improved crop yields, better fertilizer management, less pollution, and less carbon emissions. Many variables can be mapped in space and time including crop yield, terrain features/topography, organic matter content, moisture levels, nitrogen levels, pH, EC, Mg, K, and more. Farmers desire this data about their soil and now they can get it much faster. They don’t have to collect samples, send the samples out to a lab, and wait for results. Now they can have sensors and lasers attached to their equipment and get much of this info instantaneously while in the field. Importantly, they can also get many more sample points and map out the variability of soil quality in their plots.

     These precision data points that reveal soil quality in the form of nutrient levels, moisture levels, texture, and other variables can inform decisions about how much fertilizer to apply and where to apply it. It can tell them what to do and when to do it. However, some say that the promise of precision agriculture has not yet lived up to its hype.

     More precise targeting is one goal of precision ag. This includes more precise targeting of fertilizer and more precise targeting of irrigation. It also includes determining the best times to apply fertilizer and to irrigate. Thus, precise targeting and precise timing are two of the key features of precision ag. GPS/GNSS can precision guide tractors that plant seeds or apply fertilizer to optimize those processes. Another goal solved by precise targeting and timing is to decrease the amount of fertilizer and pesticides used. This can lower costs and pollution. One key to success is the sensors. Agriculture Secretary Tom Vilsack says that overfertilization is common in America’s corn fields. Better sensors and more use of sensors can lead to less fertilizer runoff.

     Farming is not what it used to be. Small farms are now the exception rather than the norm in new operations. Farming is more mechanized than it used to be and now it is more high-tech and digital. The adoption of precision ag has been slower than expected due mainly to costs and fears about it not resulting in big enough improvements to justify the additional costs. Farmers insist that the data enabled by precision ag must be actionable data, not just numbers and pictures. The seasonality of farming also slows down new technology assessment and adoption. Once the improvements from precision ag are adequately demonstrated, adoption will grow. However, now the benefits are still uncertain to many farmers, and they are worried that their investments in precision ag will not yield the results they seek. If the benefits of lower inputs and lower costs are adequately demonstrated and realized, then mass adoption is more likely.

     Bayer’s crop science division has recently been using AI to help control herbicide-resistant weeds. AI can help develop new pesticides that target specific herbicide-resistant weeds. Herbicide resistance means that more herbicides may be needed to kill weeds. That runs counter to precision ag which aims to lower inputs like fertilizer, pesticides, and diesel fuel. Some companies are focusing on robotics, developing robots that pull weeds, plant cover crops, diagnose plant infections, and gather other plant and soil data.

     John Deere introduced its autonomous tractor in 2023. However, since modern tractors have a lifespan of 25 years, many farmers will likely continue with what they have. Monocrops like corn and soybeans are more amenable to precision ag but others like fruits and vegetables are less easily integrated. There is also an agricultural labor shortage. Robots can help with that but often the crops must be altered to make robotics more effective. That is, in some cases the whole ag process has to be re-designed to better accommodate automation. Robotic cow milkers have proven to be effective, but they favor larger dairy operations due to economy of scale.

     One company that provides farming robotics, Advanced Navigation, mentions designing farms to accommodate automation, reconfiguring older equipment with automation, and developing new farming business models for equipment purchases:

 

“Our clients are companies that are designing purpose-built farm robots or retrofitting older manually operated machines to become autonomous. Farming as a service (FaaS) practices are currently becoming established as an alternative to outright equipment purchases, which allows farmers to reduce capital expenditure by contracting their required works when needed by autonomous agriculture machinery contractors. As a result, CO2 emissions are reduced due to gained efficiencies and reduced overall fuel and materials usage and will continue to shrink as more agriculture robots and autonomous ag machines become electrically powered and shared amongst farms.”

 

     A recent paper in the journal Precision Agriculture addresses fertilizer application rates to arrive at “site-specific economically optimal input rates (EOIRs) recommendations.” The study utilized machine learning models to estimate yields. Specifically, the study addressed slow-release fertilizer application rates on winter wheat crops. The results of the study showed that the selection of which machine learning algorithm to use was very important in optimizing yields and that success in improving yields can be quite variable. The study also suggested that the best algorithm to use will also vary considerably by specific farm and by specific crop. Thus, a “site-specific” approach will be needed for most applications, and this will have to be determined by gathering and analyzing data. The graph below from the paper shows some of the relationships between basal fertilization, top-dressing fertilization, and crop yields.

 

 

     Precision ag has been adopted more by larger farming operations. Its adoption among smaller farmers is lagging, particularly in tractor guidance systems. Philip Owens form the USDA’s Agriculture Research Service notes:

 

“Tractor guidance offers more spatially precise understanding of tractor operations, which lead to reduced operator fatigue, higher yield, and the ability to work longer workdays during inclement conditions. Altogether, these changes may significantly lessen a small farm's fuel, labor, repair, and maintenance costs.”

 

Source: USDA. Agriculture Research Service

He thinks there is a great opportunity for smaller farming operations to improve yields and bottom lines with precision ag and thus he promotes the use ‘big data for small-scale farmers.’ Tractor guidance is especially useful for precision planting and precision fertilizer and pesticide application. ‘Variable rate technology’ allows fertilizer or pesticide application at variable rates based on data analysis of land, soil, weeds, and crops. Combines that harvest can be outfitted with yield-monitoring systems that also provide data such as crop moisture.  

 

     Large monoculture operations are more amenable to precision ag than small diverse crop operations but as mentioned it is believed that tractor guidance can also benefit small operations since it can improve efficiencies considerably. The USDA researchers think that adoption will increase, particularly among small farms, pasture-based, and specialty crop producers.

 

References:

The Promise of Precision Agriculture Is Slowly Coming to Fruition. Eric Schmid. St. Louis Public Radio. Undark Magazine. June 18, 2024. The Promise of Precision Agriculture Is Slowly Coming to Fruition (undark.org)

Will AI conquer herbicide resistance? Ultra-precise targeting removes weeds with next to no impact on anything else. Renee Hickman, Reuters. Genetic Literacy Project. June 27, 2024. Will AI conquer herbicide resistance? Ultra-precise targeting removes weeds with next to no impact on anything else - Genetic Literacy Project

In Farming, a Constant Drive For Technology. Tom Johnson. Undark Magazine. May 4, 2022. In Farming, a Constant Drive For Technology (undark.org)

AI-based navigation systems for precision farming and agricultural robots. Advanced Navigation. 2024. Autonomous Agriculture, Precision Farming & Robotics (advancednavigation.com)

Precision agriculture. Wikipedia. Precision agriculture - Wikipedia

Can machine learning models provide accurate fertilizer recommendations? Takashi S. T. Tanaka, Gerard B. M. Heuvelink, Taro Mieno & David S. Bullock. Volume 25, pages 1839–1856. March 25, 2024. Can machine learning models provide accurate fertilizer recommendations? | Precision Agriculture (springer.com)

Benefits and Evolution of Precision Agriculture. Amanda Ashworth and Philip Owens. USDA Agricultural Research Service. Last modified July 12, 2023. Benefits and Evolution of Precision Agriculture : USDA ARS