Picking apples with a robot
According to Carnegie Mellon University (CMU), specialty crops (fruits, vegetables, horticulture and floriculture) constitute a $45 billion/year industry in the U.S. alone, of which the tree fruit and nursery industries have a farm gate value of $20 billion/year. But because the fruit are hand-picked for the most part, labor costs are exploding and represent 58% of the net value of the farm economy. So two teams of researchers at CMU are developing automated farming robotic systems to help apple and orange growers. The groups are using $10 million in grants ($6 million for apples and $4 million for oranges) from the U.S. Department of Agriculture (USDA)'s new Specialty Crop Research Initiative. But read more...
You can see above a prototype of a robot vehicle for apple growers which is currently tested at Soergel Orchards, located in Wexford, PA. Here is a link to a much larger version of this photo.
Now, let's go back to the two USDA-funded projects. The Comprehensive Automation for Specialty Crops (CASC) Program, led by Sanjiv Singh, research professor at the CMU's Robotics Institute, received a four-year, $6 million grant to develop systems for the apple industry. [Please note that I've already wrote about Singh last month in Robot helicopters flying low among obstacles.]
And the Integrated Automation for Sustainable Specialty Crop Farming Project, led by Tony Stentz and Herman Herman of the Robotics Institute's National Robotics Engineering Consortium (NREC), received a three-year, $4 million grant to develop systems for the citrus industry. Both project grants will be matched dollar for dollar by industry, state governments and other funding sources.
Now, let's look at the CASC site. The Singh's team is actively working at Soergel Orchards. Here is a link to the visit the team made on November 14, 2008. "We did our first autonomous run at Soergel Orchards using an electric utility vehicle that has onboard computing and sensing. The path followed was 1.3 km long using GPS waypoints at speeds between 2 and 3 m/s. The path was four 150 m legs in rows spaced about 26 feet between centers." You can see a short video from the above link and see more pictures from this image gallery.
The team returned to Soergel Orchards on November 24, 2008 to acquire range data using laser scanners . Of course, the researchers took a bunch of other pictures.
For more information, you also can read Orchard robotics makes one juicy project (David Templeton, Pittsburgh Post-Gazette, December 3, 2008). Here are the last paragraphs of this article. "Mr. Soergel said 'the most exciting part is keeping prices as low as possible and cutting costs' at the orchard his family has operated since 1850. One priority is giving an early alert to growers when insects arrive or disease shows up. For now, Dr. Singh's team is developing a way for their robotic vehicle to understand where it's situated in the orchard. Larger orchards already use moving platforms that proceed slowly, tree to tree, and can be raised or lowered depending on the fruit's location. An immediate goal is a platform that travels without a driver. 'We're taking baby steps,' Dr. Singh said. Or, as he notes with a wry smile, 'we're picking the low-hanging fruit.'"
Sources: Carnegie Mellon University Robotics Institute news release, November 19, 2008; and various websites
You'll find related stories by following the links below.