Farm Progress is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Serving: East
Corn+Soybean Digest

How Far Off Are Farm Robots?

Are you ready to give up driving your tractor? Engineers are perfecting driverless autonomous, or independent, tractors and robots. They make choices on the go based on sensor-generated readings of field conditions.

“Farming will never be the same again,” says Bill Stout, Texas A&M ag engineering professor emeritus. “Technically, we can put a man on the moon, use driverless tractors, but economics and liability concerns will govern the widespread use of driverless tractors. Robotic equipment is not a new subject. I saw a robotic rice harvester in Japan 10-15 years ago, which followed uncut rows and turned automatically at the row's end. New Holland had a driverless forage harvester many years ago.”

The next step in precision agriculture invokes this technology to treat crops and soils selectively according to their needs by small autonomous machines, says Simon Blackmore, visiting ag engineer from the University of Thessaly, Greece, and at Bristol Robotics Laboratory, England, and managing director of Unibots. This next step employs phytotechnology: reducing the scale down to an individual plant.

Blackmore sees smart machines tilling, weeding, spraying and harvesting crops. “In 10-15 years you may see smaller ‘slave’ tractors that operate semi-autonomously alongside manned farm tractors. These smaller vehicles create less soil compaction, can operate in poor weather and soil conditions and are more ‘scalable’ to fields of varying size,” he says.

Blackmore envisions less intrusive shallow plowing machinery that minimizes energy use and soil compaction. “Most of the energy used in traditional cultivation is to repair the compaction from large tractors,” he says. “Micro tillage (a few cubic centimeters) will focus on the position where the seed will go. Perhaps we will develop smaller, less intrusive machinery operating in marginal weather and soil conditions that are optimal for the crop,” Blackmore says.

“Precision geospatial mapping of each seed's placement is relatively simple as RTK (real-time kinematic) GPS is fitted to the seeder, and infrared sensors mounted below the seed chute,” Blackmore says. “As the seed drops, it cuts the infrared beam and triggers a data logger that records the position and orientation of the seeder. A simple kinematic model can then calculate the actual seed position, and seed coordinates can be used to target subsequent weeding and harvest operations.”

Precision technologies make possible, for example, hexagonal seed placement, which maximizes the distance between seeds in all directions for maximum access to nutrients and sunlight. (Hexagonal spacing still has rows — but at a different angle.)

“Robotics are most valuable when they can perform field operations more accurately than what is humanly possible,” says Clay Mitchell, Mitchell Farms, Geneseo, IA. He is known nationally as a pioneer in exploiting automation technologies for agronomic gains. In the past seven years, he's used automation technologies to improve the placement of seed, chemicals, fertilizers and to reduce fuel use. “I believe that improving yield is more important than reducing labor in the adoption of robotics,” he says.

Crop scouting can also benefit from robotic and sensing technology. Imagine a high-clearance autonomous platform guided by GPS with a range of sensors to measure crop nutrient status and stress, identify weed species and calculate weed density. One such vehicle, the ISAAC2, was developed by a student team at Hohenheim University (

Robotic weed mapping records the position and density (biomass) of different weed species using machine vision technologies, Blackmore says. One method classifies at least 20 weed species by the shape of their outline. The weed map then becomes the basis of a treatment program.

Adaptive weed control technology, developed by Tony Grift, ag engineer at the University of Illinois, uses laser beams to detect weeds. Along with colleagues from Japan's National Agricultural Research Center, he is developing platoons of Flock-Bot robots. Groups of five to 10 such robots can patrol fields night and day to locate, identify and eliminate weeds. Costing only $3,000-5,000 each, the robots will be guided by RTK-GPS (an additional $15,000), remote control or by navigating relative to crop rows.

Their batteries will be boosted by solar panels, and/or by burning crop waste to make steam power. A web cam will send images to a database to compare with weed identification parameters.

Weeds will be controlled either mechanically by a precision harrow mounted on the back of the robot, a saw blade sprayed with pinpoint herbicide micro-sprays or burned with highly focused reflected sunlight. The harrow option involves a smart sensor directing a retractable arm to move in and out between plants to cutoff a verified weed.

Robotic harvesting could be selective, identifying and harvesting only grains or fruits meeting certain quality thresholds, Blackmore says. By 2025, Blackmore envisions a group of 10 autonomous harvesters, each with a 3-ft. head stripping the ears directly from the stalks. The grain and chaff could be brought back to the farm for threshing with a stationary threshing machine similar to the hand harvesting system used many years ago.

“Ten of these small autonomous harvesters have the same cutting width as a modern combine, but they could carry out selective harvesting,” Blackmore says.

Consider this. In order to cross a field, especially an uneven one, driverless tractors must be able to see in three-dimension. As with humans, a tractor's automatic guidance system needs two eyes, or a stereo camera, to perceive depth.

Qin Zhang, ag engineer at the University of Illinois, uses stereo cameras on tractors to assess the terrain and steer the machine — without anyone behind the wheel. This work culminated three years ago when he tested guidance technology enabling a common model of farm tractor to guide itself across a farm field.

“With curved rows and straight rows, we successfully achieved a speed of 8 mph without any problem,” says Zhang. He expects to bring cruising speed up to 12 mph before long.

Another promising type of sensing technology is laser scanners, says Scott Shearer, ag engineer at the University of Kentucky. Lasers can profile the ground and any obstructions quickly, measure distances and generate images of terrain in milliseconds.

In research conducted for citrus growers, laser radar (ladar) sensors provide the “vision” for identifying the end of the row, locating the row to be turned into and positioning the tractor relative to the citrus trees during turning.

“We'll see more sensing capabilities that more accurately replace what the farmer verifies over his shoulder,” Shearer says. “For example, a combine will monitor the quality and moisture of grain entering the tank, along with a whole host of factors like the strength of the corn cob. Planters will verify that seed is dropping and in contact with the soil, and that the furrow is properly covered.”

Sensors have a tall order in judging what drivers do almost automatically, such as turning at the end of the row. Engineers have worked long and hard to duplicate the process in robotic technology.

Some of these technologies were first developed by the defense industry. Already, the military deploys robotic scouts in Afghanistan, says Bingcheng Ni, Case New Holland project engineer, innovation.

To encourage continued innovation in military robotics, DARPA (Defense Advanced Research Projects Agency), which is credited with conceiving the Internet, awards$2 million to the first autonomous vehicle that completes a 142-mile Mohave Desert route of GPS coordinates. “This would have been unimaginable five years ago,” Case's Ni says. “The first year, no one finished more than 10% of the route.”

Eighteen months later, four autonomous vehicles successfully completed a 132-mile desert route under the required 10-hour limit. This November the competition will move to the city, requiring autonomous vehicles to navigate and operate in traffic.

On the farm, some growers still do not relish the idea of leaving their tractor cabs, regardless of the prospect of improved technology. “The barrier is human nature,” says University of Illinois' Grift in explaining the advantage of reduced rollover accidents in unmanned tractors. “A farmer says ‘That's what I do — drive a tractor.’”

Perhaps this hurdle of tradition, along with safety and liability concerns, is what prevents ag equipment manufacturers from introducing autonomous vehicles in the near future. “Driverless tractors are less likely to appear in John Deere's lineup in the near term,” says Barry Nelson, Deere public relations manager. “Because of safety considerations and cost requirements, this is not a viable solution for the majority of our customers.”

“We will continue research and development in new technology to build and design tractors and vehicles that provide the best value and productivity for our customers,” he says.

How Soon?

When might you see robotic tractor demos at a farm show or in the neighbor's field? Part of that hinges, some say, on what happens with U.S. immigration reform. A scarcity of farm labor could hasten acceptance of robotic tractor technologies, says Scott Shearer, ag engineer at the University of Kentucky. “The qualified labor pool is dwindling. You don't put just anybody in the seat of a combine today,” he says

A second reason might be an increase in capacity, says Jan Willem Hofstee, Farm Technology Group, Wageningen University, The Netherlands. “Using state-of-the-art technology for guiding results in a capacity increases because for a human driver it is very difficult to oversee a header of a 33-ft. combine (or even more). State-of-the-art technology can steer very accurately and use almost the full width of the header while a human driver may miss by 2-3 ft.

“The next five years may see auto-steering and automated headland turning evolve to where tractor drivers playa more supervisory role. You will also likely see small, unmanned airplanes shooting photos for scouting imagery, mapped robotic seeding and robotic weeding. In 5-15 years, Midwestern farmers will see tractors that can do the work autonomously without a driver in the driver's seat,” says Hofstee.

TAGS: Equipment
Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.