Contributed by Grace Curtis | Junior | Journalism
The Cal Poly Strawberry Center, an organization formed by a partnership between California Polytechnic State University and The California Strawberry Commission, is the only organization in the U.S. dedicated solely to strawberry research and education. It provides students access to some of the best Learn By Doing opportunities Cal Poly has to offer.
The Center has a unique position on the Central Coast of California, a location that is home to 90% of the country’s strawberry production. Its goal is to increase sustainability in the strawberry industry through research, and recently Center researchers have launched a new project to improve the use of automation in harvesting strawberries.
Dr. Mojtaba Ahmadi, Senior Production Automation Engineer at the Center, has been leading a time-lapse photography project with several interns where they track the growth of several strawberry plants by taking pictures of them every day. After they’ve captured a three month long time-lapse of photos, they will use the data from the images to create simulations where robots can be tested and later implemented into the fields.
Why is automation needed in strawberry fields?
Strawberry plants have growth patterns that pose a unique challenge for farmers. The plants have structures called “runners” that they use to regenerate themselves. These long stem-like organs sprout from the main plant, extend over to another spot in the soil, and develop roots in a new area, effectively creating a whole new plant.
While a strawberry plant multiplying itself sounds like it could be beneficial for mass growth, it actually reduces crop production. Not only is it important for farmers to prevent this multiplication from happening to maintain organization in the fields, but the formation of runners uses up nutrients from the plant that could otherwise be used to grow new leaves and therefore fruit as well.
Right now, the Cal Poly Strawberry Center and California Strawberry Commission use manual labor to cut off all these runners when they form. Creating a robot that could cut these shoots automatically would save both time and money and would help the industry move towards a less labor-intensive system.
Collecting data from the time-lapse project
Ahmadi says that when creating software to test the algorithms the robots will use in the fields, it is highly important to use data from real plants rather than randomly generated numbers. This project will help his team gather data from the growth of real plants, tracking the formation of several organs such as leaves, flowers, fruit, and runners.
“They need to test the robot somehow. They can’t just take it into a commercial field,” said Ahmadi. “It would be nice to have such a model so that they can virtually test a robot, see how the algorithm is working, whether they're making right decisions, the right moves or not, and then fix those bugs.”
The project is just over a month in, and Ahmadi and his team hope to continue the project for at least another couple of months. Soon, they will start the data processing phase of the project, and he is looking for another intern to help out with that portion.