Imagine a future where you wake up in the morning and a color-coded map of your orchard is shown on a touch-screen display. You sit in front of the screen with the coffee still hot and start querying the display for the status of several important conditions you need to monitor. What’s the moisture status in that pocket of thirty trees in zone 4 of block 21 that were water-stressed just last week? Are codling moths still trying to make an incursion into block 45, or has the automatic monitoring trap shown a decline in activity? And what about that one tree in block 12, the one bordering the game reserve, where yesterday the display alerted you to the possible onset of fire blight?
You finish your coffee and turn off the display, happy to see that the precision irrigation system did a good job irrigating those specific 30 trees more than the adjoining ones, and that there are no new signs of pest activity anywhere in the orchard. You also make it your first priority to drive to block 12 and check those troubling signs of possible fire blight.
As you step out to the orchard, you see the driverless electric vehicle docking autonomously at the battery recharge station. You look forward to the crop load report that will be available at 10 a.m., after all the 100,000 pictures taken by the vehicle during the night are processed. You also inform your field manager that today the vehicle will execute an automatic canopy volume measurement run, the results of which may indicate the need for some summer pruning. He in turn alerts you that his crew will be testing the new bin filling system, which competing farms have been using successfully to increase harvest efficiency. You can’t help but think of how much harder and costlier it was to manage your operations back in 2010, when you first experimented with some of these technologies, and wonder what other technologies will be available 10 years down the road.
While this is certainly a visionary scenario, it is one we are right now hard at work to make true. Comprehensive Automation for Specialty Crops (CASC) is a 4-year project funded by USDA and the industry to develop, test, deploy, and commercialize various systems to improve orchard management, streamline major operations such as harvest, and reduce growers’ costs.
Fulfilling Growers’ Needs
While initially focused on the apple and nursery industries, most of the underlying technologies are applicable to other tree fruit crops. Our work is driven by actual growers’ needs, and project members interact with stakeholders on a constant basis. CASC encompasses three major efforts:
1. Information Management: Methods and devices to detect water- and nutrient-induced plant stress, detect plant diseases such as fire blight, count and classify insects such as codling and oriental fruit moth, measure crop load, and classify nursery trees by caliper diameter; and systems to organize all this high-resolution, high-volume information in a way that it is easy to manipulate, understand, and apply to operations planning and farm management.
2. Automation: Machines for automatically deploying sensors; for automating farm operations such as precision spraying and mowing; and for increasing efficiency of labor intensive tasks such as harvesting and thinning.
3. Technology adoption: Work aimed at understanding the socio-economic barriers to technology adoption and how to overcome them; at quantifying the benefits of individual technologies and systems and calculating their return on investment rate; and at educating growers on the use of these new technologies and systems.
CASC is led by Prof. Sanjiv Singh from Carnegie Mellon University’s Robotics Institute in Pittsburgh, PA, and includes more than 50 engineers, scientists, and economists from Penn State University, Washington State University, Oregon State University, Purdue University, USDA Agricultural Research Service, Vision Robotics, IONco, Toro, and Trimble. The project effectively started in September 2008 and will continue until August 2012. In the past 10 months the CASC team has produced and field-tested prototypes of several systems that will mature in coming years to commercial products:
• Cameras and software that automatically detect the onset of fire blight.
• A digital trap that automatically counts insect-size objects that fall through it. In the future, it will be able to distinguish target insects from other insects or objects.
• A laser caliper that automatically measures nursery tree diameter.
• Cameras and software that automatically determine crop load.
• Fruit handling and bin filling mechanisms that increase worker productivity and reduce bruising.
• Software that organizes fruit, tree, and block information in graphical formats to facilitate farm management and increase efficiency.
• Automated driverless vehicles that drive safely along the rows deploying sensors and carrying farm implements such as precision spraying systems and mowers.
Commercializing New Products
In addition to developing these systems, our team ran extensive surveys to understand the socio-economic barriers to technology adoption and how to overcome them when developing or later commercializing new products. We will continue gathering input from stakeholders for the duration of the project to map the evolving needs of growers and adapt our offerings accordingly. We have also started quantifying the economic aspects of each system we are working on, to show they not only solve a well-defined need but also are economically sound from the point of view of the return on investment.
More information about CASC, including pictures and videos of recent field trips, can be found at http://www.fieldrobotics.org/casc.