Oct 3, 2013Apple harvesting sorting system makes debut
Researchers at Michigan State University (MSU) have unveiled their first prototype of an in-orchard harvesting aid/sorting machine that could prove to be significantly lower in cost than other existing systems.
Renfu Lu, research leader for the USDA/ARS Sugarbeat and Bean Research Unit at MSU, said the machine has been in development for about four years with support from the Michigan Apple Committee.
“This machine allows growers to achieve cost savings post harvest, reduce post harvest disease and pest problems and enhance product choice-ability – those are the three main goals,” Lu said at a Sept. 12 demonstration of the machine at the farm of Brett Anderson in Sparta, Mich.
“This system can help growers improve harvest efficiency and worker safety and in the meantime, it will sort apples into two or three quality grades (cull, processing and fresh) in the orchard, so that growers do not need to send inferior fruit to the packinghouse,” Lu said.
Currently, Lu stated, both defective and sound apple fruits are not sorted at harvest, but are combined, causing significant storage losses due to the defective fruits being susceptible to pest and disease infestation. The result is costly, postharvest handling to remove unmarketable fruits.
“We developed an in-orchard mobile system that automatically sorts and grades apples into culls (defective), processing, and fresh-market quality fruits by measuring fruit color, size, shape and weight using color-imaging, machine-vision technology,” Lu indicated. “This system incorporates harvest aid functions to reduce safety hazards for fruit harvesters. This technology will enable apple growers to separate or leave defective fruit in the orchard, resulting in less postharvest disease/pest problems and lowering postharvest storage and packing costs, and will assure a better fruit quality inventory at the warehouse.
Lu indicated the mobile system is utilizing an existing commercial apple harvest trailer and a tractor to reduce the overall cost. It uses a low cost, computerized color imaging technology, coupled with several novel designs in fruit handling, grading, and bin filling, to sort and grade apples at a speed of 6-8 fruits per second.
“Hence, the overall cost for the machine will be much lower than other existing systems that are either under development or being commercialized recently,” Lu said. The mobile system can sort apples into two (fresh and cull) or three (fresh, processing and cull) grades, depending on grower requirements. The system incorporates flexible picking conveyors (or arms), allowing six to eight people to pick apples from the ground and the platform.
“Cost, use and functionality are the primary concern in the design of the system,” Lu said.
The fully automated apparatus uses a computer visioning system to perform the sorting. A camera takes 10 to 15 images of each apple, and the computer analyzes what grade and bin the apple should be placed.
“We have three bin fields,” he explained. “Each has a microcomputer chip. The system can be programmed to control the bin field, and decides at what control weight it wants to drop the apple. A computer touch screen is used to change which bin you want to fill.
“Compared to other systems we are already familiar with, we’ve designed our system on a commercial trailer to really reduce the cost and can adapt our system to already existing harvesting equipment that growers already have.”
He said the sorting effort could be put in practice in the field, or in a farmer’s own packing area when they have apples that have been recently harvested.
Lu started the project after meeting with growers who were seeking a method to sort apples in the field so they can curtail the shipment of low quality fruit into the packinghouse and reduce the cost of sorting and packing.
“To do that, you have to own a machine, which also has a cost,” Lu said. “So we did an economic analysis that found in order to do so you have to have a cost-effective machine. We believe we now have the machine.”
He said the bulk of the costs of the rig come from the computer and software, camera and conveyor belt motor, which has been adopted from a typical vehicle windshield wiper application.
Lu said the machine is not yet under patent and a commercial vendor is being sought to continue the development of the project.