American apple growers can't predict the yield of each year's crop, nor the size of fruits that make up annual harvests.
"Orchard agriculture is the exact opposite of a modern factory," said Sanjiv Singh, a research professor at Carnegie Mellon University's Robotics Institute. "In manufacturing, managers can decide how many of something they produce and when."
But growing specialty crops -- apples and oranges, for example -- relies on people walking among rows of trees, examining fruit health, checking for bugs and noting branches that need to be thinned or pruned.
"What we're trying to do is increase the efficiency of producing apples," Singh said last week, standing at the edge of one of the Soergel Orchards in Franklin Park.
A robot sitting about 10 yards away might help them do that, say Singh and a team of scientists.
The robot, a red Toro four-wheel-drive vehicle, is part of a four- year, $6 million project funded by the U.S. Department of Agriculture that began last year, said Daniel Schmoldt, a USDA official based in Washington.
Using lasers that sweep a 200-degree field of vision and computer programs that locate the middle of the tree rows, the robot maneuvers inside the orchard at about 3 mph, said Bradley Hamner, a CMU Robotics Institute researcher.
A team including scientists from Penn State, Purdue, Oregon State and Washington State universities expects that type of automation will transform specialty-crop farming into a less labor-intensive and more profitable venture, said Clark Seavert, an Oregon State agricultural economics professor. …