Jul 30, 2015Ground truthing technology utilizes sensors to determine crop trends
Agriculture today has transformed into a high-tech enterprise that most 20th century farmers might barely recognize.
Alex Thomasson, professor of biological and agricultural engineering at Texas A&M University, said it was only 100 years ago that farming in the U.S. transitioned from animal power to combustion engines.
“Over the past 20 years, the global positioning system (GPS), electronic sensors and other new tools have moved even further into a technological wonderland,” Thomasson wrote in an article published in The Conversation, an open compilation of views expressed by the academic research community.
“A modern large tractor’s enclosed cabin includes computer displays indicating machine performances, field position and operating characteristics of attached machinery like seed planters,” Thomasson stated. “And as amazing as today’s technologies are, they’re just the beginning. Self-driving machinery and flying robots able to automatically survey and treat crops will become commonplace on farms that practice what’s come to be called precision agriculture.
“The ultimate purpose of this high-tech gadgetry is optimization, from both an economic and environmental standpoint,” Thomasson wrote. “We only want to apply the optimal amount of any input (water, fertilizer, pesticide, fuel, labor) when and where it’s needed to efficiently produce high crop yields.”
Thomasson talks about using a GPS receiver to locate preselected field positions to collect soil samples. Then a lab analyzes the samples and creates a fertility map in a geological information system.
“That’s essentially a computer database program adept at dealing with geographic data and mapping,” he wrote. “Using the map, a farmer can then prescribe the amount of fertilizer for each field location that was sampled. Variable-rate technology (VRT) fertilizer applicators dispense just exactly the amount required across the field.”
The three things that are essential to make precision agriculture successful, according to Thomasson, are site-specific information, the ability to understand and make decisions based on that information, and the physical tools to apply the management decisions.
He said computer models that mathematically and statistically analyze relationships between variables like soil fertility and the yield of the crop often aid decision-making. The GPS-enabled VRT fertilizer applicator is an example of a physical tool that automatically adjusts its rate as appropriate for each field position.
Other examples of precision agriculture involve varying the rate of planting seeds in the field according to soil type and using sensors to identify the presence of weeds, diseases or insects so that pesticides can be applied only where needed.
Agriculture is becoming comfortable with the use of remote sensing to acquire information about an object or phenomenon without making physical contact with the object.
Data drives tree fruit efforts
“Growers are interested in collecting as much data as possible, and aerial imagery is the foundation of that,” said Herman Thoennissen, owner of HTG International in Kennewick, Washington.
“Infrared brings out detailed colors and helps identify problems,” Thoennissen said at the most recent Washington State Horticultural Association annual meeting. “Wouldn’t you like to see this so you can take corrective action? But you have to hire people who can use it correctly. Don’t focus on the devices. Focus on the people using the devices. You have to empower the mobile workforce.
“I am excited as ever – I wish I was 26 again,” Thoennissen said of the evolving technology.
“It can help identify canopy density in fruit trees over time,” he said. “It’s been done in agronomy for some time. It will become the foundation for the (tree fruit) industry in the near future. We are on a very steep learning curve.”
Aerial sensor technologies to detect and classify objects on Earth by means of propagated signals has taken the form of “ground truthing,” which refers to information collected on location.
A common definition of ground truthing is it allows image data to be related to real features and material on the ground. The collection of ground-truth data enables calibration of remote-sensing data, and aids in the interpretation and analysis of what is being sensed.
Ground truthing can help fully identify objects in satellite photos.
Thomasson and Thoennissen both touted the use of unmanned aerial vehicles (UAVs, or drones) that can collect highly detailed images of crop and field characteristics.
“These images, whether analyzed visually or by computer, show differences in the amount of reflected light that can be related to plant health or soil type, for example,” Thomasson said. Once disease extent is identified in a field, future treatments can be applied only where the disease exists.
“Advantages of UAVs include relatively low cost per flight and high image detail; the legal framework for their use remains under development,” Thomasson said.
Sensors are in use in other ways. Thomasson said fully autonomous or robotic field machines have begun to be employed in small-scale, high profit-margin agriculture such as wine grapes, nursery plants and some fruits and vegetables.
“Autonomous machines can replace people performing tasks, such as hand-harvesting vegetables,” he said. “They use sensor technologies, including machine vision that can detect things like location and size of stalks and leaves to inform their mechanical processes.”
He noted Japan is a trendsetter in sensor technology.
“Typically, agriculture is performed on smaller fields and plots there, and the country is an innovator in robotics,” he said. “But autonomous machines are becoming more evident in the U.S., particularly in California, where much of the country’s specialty crops are grown.”
Thomasson believes the development of flying robots gives rise to the possibility that most field-crop scouting currently done by humans could be replaced by UAVs with machine vision and hand-like grippers.
“Many scouting tasks, such as for insect pests, require someone to walk to distant locations in a field, grasp plant leaves on representative plants and turn them over to see the presence of insects. Researchers are developing technologies to enable such flying robots to do this without human involvement.”
Technology helps breeders
Another breakthrough is taking place with high-throughput plant phenotyping (HTPP) – “a technology at the intersection of genetics, sensors and robotics,” Thomasson wrote.
“It is used to develop new varieties or ‘lines’ of a crop to improve characteristics such as nutritive content and drought and pest tolerance.”
He said HTPP employs multiple sensors to measure important physical characteristics of plants such as height, leaf number, size, shape, angle, color or wilting.
“The sensor combinations can very quickly measure phenotypic trains on thousands of plants on a regular basis, enabling breeders and geneticists to decide which varieties to include or exclude from further testing, tremendously speeding up further research to improve crops,” Thomasson said.
“Don’t be surprised if 10 years from now you drive down a rural highway and see a very small helicopter flying over a field, stopping to descend into the crop, use robotic grippers to manipulate leaves, cameras and machine vision to look for insects and then rise back above the crop canopy and head toward its next scouting location – all with nary a human being in sight.”