“Pollination is one of the important operations in orchards every year,” Karkee said. “Crops like apples, cherries, pears, and many more require pollination.”
Currently, crops are pollinated mainly by honey bee hives, which beekeepers move around to different orchards and crops. However, pollinator populations are decreasing and robotic pollination could be a more stable and efficient process.
The project will involve three steps:
Camera and machine learning systems. The research team must teach machines to detect and locate flower blossoms on trees in an orchard to target what needs to be pollinated.
Look at models for blossom development. Researchers need to determine when is the best stage to pollinate a given crop and a given location.
Develop a robotic hand and arm to spread pollen that will work with step one and will be tested in a lab and in the field.
Luckily, the team won’t be starting from scratch.
“We’ve been working for almost a decade on robotics in agriculture,” Karkee said. “For example, we have robotic systems that detect flowers for a robotic thinning system. We just need to adapt that to work for this project.”
The robotic pollinators also have hands and arms that can pick crops like apples, which they plan to adapt for pollinating blossoms.
The overall goal is to demonstrate capability for robotics to do this work, not build a full-scale machine, Karkee said.
“This is an important project, and a lot of work, but we’re confident we’ll be able to put together all of these different pieces into a viable prototype that could be a huge help for the agriculture industry in the future,” he said.
The project includes horticulturists from both WSU and Penn State, as well as engineering faculty from both universities. In addition to Karkee, the WSU team includes Changki Mo (Mechanical Engineering), Matt Whiting (Horticulture), and Qin Zhang (BSE). The Penn State team includes Jim Schupp (Horticulture) and Long He (Biological Systems Engineering).