2705:
Automation for Specialty Crops: A Comprehensive Strategy, Current Results, and Future Goals

Sunday, July 26, 2009
Illinois/Missouri/Meramec (Millennium Hotel St. Louis)
Sanjiv Singh, Research, Professor , Robotics Institute, Carnegie Mellon University, Pittsburgh, PA
Tara Baugher , Penn State University, Gettysburg, PA
Marcel Bergerman , Robotics Institute, Carnegie Mellon University, Pittsburgh, PA
Ben Grocholsky , Carnegie Mellon University, Pittsburgh, PA
Jay Harper , Penn State University, University Park, PA
Gwen-Alyn Hoheisel, Area, Extension, Educator , Washington State University, Prosser, WA
Larry Hull, Professor , Penn State University, Biglerville, PA
Vincent Jones , Washington State University, Wenatchee, WA
George Kantor , Robotics Institute, Carnegie Mellon University, Pittsburgh, PA
Harvey Koselka , Vision Robotics, San Diego, CA
Karen Lewis , Washington State University, Ephrata, WA
William Messner , Carnegie Mellon University, Pittsburgh, PA
Henry Ngugi, Assistant, Professor , Penn State University, Biglerville, PA
James Owen Jr. , Oregon State University, Aurora, OR
Johnny Park, Principal, Research, Scientist , Purdue University, West Lafayette, IN
Clark Seavert , Oregon State University, Aurora, OR
Specialty crops constitute a $45 billion/year industry that is steadily growing. As opposed to broadland crops (such as wheat and soybean), they are characterized by the need for intensive cultivation. This requires a skilled, efficient, cost-effective labor force. This diminishing labor force is limiting economic returns and adversely affecting the cost of production resulting in a labor crisis. In addition, an increasing consumer demand for a safe, affordable, traceable, and high quality food supply and the need to minimize the environmental footprint represent key challenges for specialty crop sustainability in the United States. Recently, our team started a 4-year effort to develop a comprehensive automation strategy for tree fruit production. Our approach is based on a triad that spans the entire production spectrum from nursery production to fruit harvest that includes: (1) Information, mobility, and manipulation technologies (reconfigurable mobility, positioning, information management, and augmented fruit harvesting). Reconfigurable mobility creates low-cost, robust moving platforms that can be tasked based on specific needs of the growers. Accurate positioning provides the capability of georeferencing specific observations even when GPS is unavailable. Information management deals with the multi-scale, spatio-temporal data that must be integrated into a common geographic information system. Finally, assisted harvest focuses on increasing efficiency of fruit picking and reducing fruit damage in the field. (2) Plant science technologies (plant stress and disease detection, monitoring insect populations and infestations, automated caliper measurement, and autonomous crop load scouting). Biotic stress such as pest damage account for 10-30% of yield losses while abiotic stress, such as nutrient and poor water management, account for 70-90% of the yield reductions. Agricultural progress has primarily identified factors that limit productivity and provide either genetic improvement or cultural strategies to mitigate the stresses. We focus on tools to detect plant stress and crop load and integrate this information into management techniques that will increase productivity and quality of the final product. (3) Outreach and commercialization. Research shows that it takes roughly eight years for new technology to be incorporated by early adopters, and as long as 15 years for full industry adoption. We intend to reduce this amount of time significantly by: (a) proactively fielding the technologies developed in real apple orchards and tree nurseries in Pennsylvania, Oregon, and Washington, and (b) partnering with commercial technology providers from the outset, so technology transfer happens as a natural by-product of the effort.