- SPECIAL REPORTS
That experience prompted the inner farm boy in me (I come from a long line of farmers on my mother’s side) to ponder the future of agriculture. In particular, what role will robotics play in farming and horticulture tomorrow?
Autonomous tractors and smart implements are not new. In fact, way back in 1933, International Harvester Co. displayed a driverless, remote-controlled tractor at the Century of Progress world’s fair in Chicago. More recently, engineers at Deere, New Holland and other companies have created autonomous vehicles equipped with state-of-the-art computers, lasers, sensors, GPS systems and machine vision technology. Robotic milking machines have been on the market since the early 1990s.
However, recent advances in robot reliability and deployment cost, coupled with changing labor demographics, have sparked a new interest in automated agricultural equipment. Engineers around the world are busy developing a wide variety of robotic devices for seeding, tilling, weeding, spraying and harvesting applications.
For instance, a robotic planter has been developed by engineers at the University of New South Wales. It automatically corrects lateral errors in seeding while improving crop rows. The device works in conjunction with an autonomous tractor. The Australian engineers have also invented a robotic weeder that identifies pesky plants and destroys them with an electric charge.
Engineers at the UK’s University of Warwick have created a robotic mushroom picker. It uses a charge-coupled camera to pinpoint the coordinates of each mushroom and select the exact size required. A laser determines each mushroom’s height. The actual picking is done by a robotic arm that’s equipped with three vacuum grippers.
Also in England, the National Physical Laboratory (NPL) is working with FANUC Robotics, KMS Projects and Vegetable Harvesting Systems to create an intelligent machine for harvesting cauliflower, lettuce and other big-leaf crops. They’re using radio frequencies, microwaves, terahertz and infrared technologies to probe beneath the leafy layers of a crop, identify different materials and enable precise size identification.
“These four parts of the electromagnetic spectrum all have potential to safely penetrate the crop layers and identify the size of the harvestable material for a relatively low cost,” says Richard Dudley, NPL’s principal research scientist. Traditionally, human eyes have difficulty determining whether leafy crops are ripe for picking As a result, Dudley claims that annual waste for certain crops can be up to 60 percent.
In Holland, Wageningen University sponsors an annual Field Robot Event. The contest allows university teams to test and benchmark their latest agrobot designs. Winners of the 2009 event were EasyWheels, from the Helsinki University of Technology; EyeSonic, a three-wheeled device from Wageningen University; and Helios, a four-wheel vehicle from Braunschweig Technical University.
Engineers in Japan are also tackling the challenges of robotic farming. For instance, Okayama University’s Laboratory of Agricultural Systems Engineering has developed different types of mobile platforms equipped with articulated robots for fruit and vegetable harvesting applications. Robots use vacuum grippers to pluck cherries, cucumbers, strawberries and tomatoes. A 2D vision system allows the devices to detect and select green fruit instead of green leaves and stems.
Closer to home, engineers at Ohio State University have used vision-guided, six-axis robots from Motoman Inc. to harvest tomatoes using four-fingered grippers. And, a company in California called Vision Robotics Corp. has developed robotic equipment for harvesting apples and oranges. Vision systems are first used to scan and identify ripe fruit. Cameras placed at the end of long booms create a 3D image of an entire tree. Then, a series of long reticulating arms equipped with grippers are deployed to pick ripe fruit.
Several other projects currently underway in the United States are aimed at automating vineyards, apple orchards and orange orchards. The goal is to develop a fleet of robotic pruners, sprayers and fruit pickers. Engineers are Carnegie Mellon University’s Robotics Institute recently received a $10 million grant from the U.S. Department of Agriculture to create automated equipment that can lower production costs.
“We are taking automation to a level never before demonstrated in an agricultural setting,” says Herman Herman, principal commercialization specialist at the Robotic Institute’s National Robotics Engineering Center. “This will provide an early look at how the automated farm may someday operate.
“Harvesting remains one of the most labor-intensive operations at orchards, but it is very challenging to automate because of demanding handling and cost requirements,” Herman points out. However, he claims that the goal of the Integrated Automation for Sustainable Specialty Crop Farming Project is “aiding human harvesters rather than replacing them.”
Engineers from Cornell University, Deere & Co. and the University of Florida are also involved in the three-year research effort, which hopes to “design a cost-effective mechanical harvester, including specific perception, planning and picking components” by 2011. They expect to achieve a 75 percent reduction in tractor-related labor costs. Testing of autonomous tractors, sprayers, automated harvesters and other equipment is taking place at Southern Gardens Citrus in Florida.