Florida undergrads plan to upend hydroponics with machine learning, IoT – and Jiffy growing media. Munir Hafeel, Andres Palacios, Daniel Katsale, and Kelan Zielinski, all undergrads at the University of Miami, built a hydroponic basil farm that runs itself using machine learning and the internet of things (IoT).
Munir, who led the project, got into hydroponics and computer engineering as a teen. He eventually won a scholarship to study the latter at the University of Miami, but growing produce always remained a passion. So he decided to try to improve CEA yields using computer engineering for his senior design project.
Nine million data points per month
Munir’s team built a hydroponic system from scratch in a fifty-square-foot room at the university. They included sensors which read nine million data points a month on factors like pH, temperature, humidity, and nutrient levels. The data is continually analyzed using machine learning to optimize growing conditions using various algorithms.
“If the sensors detect too few nutrients in the water, an algorithm automatically tells the peristaltic pumps to release more,” says Munir. “We designed similar systems to regulate pH and dissolved oxygen levels.”
Although the project’s space is small, the team wanted to create a scalable concept. They worked with Microsoft to develop scalable software hosted on the cloud and create a plug-and-play hydroponic system that would be capable of taking care of everything for commercial growers.
More control equals less risk for growers
Above all, the team wanted to add an additional level of control to hydroponics by making the data available remotely. And to reduce hydroponics’ uncertainties in order to make food production more accessible and sustainable. “Computer engineering helps minimize resource use and maximize harvests,” he explains, “which can make farming less risky and stressful for growers.”
Deciding which growing media to use was easy. “I think Jiffy Pellets go a long way, especially when it comes to germinating and maintaining the plant. So I was really happy to work with Jiffy.”
Learning from Jiffy experts
In fact, Munir had an additional reason to choose Jiffy: He initially learned his hydroponics skills from the experts at Jiffy’s production plant in Sri Lanka, where Jiffy Growbags and coir pellets are produced, and where Munir is from. “I would never have known about hydroponics if it wasn't for Jiffy.”
“I tried sowing some tomato seeds when I was 14, but they barely germinated. When I came across an ad for Jiffy I reached out for help and the plants finally started growing. I ended up building a rooftop greenhouse containing hundreds of growbags complete with simple sensors. I wrote about my experience setting this up for my college essay, which helped me gain admission to the University of Miami.”
For their senior project, Munir, Andres, Daniel, and Kelan transplanted the seedlings two weeks after sowing and the basil was ready to harvest four weeks later, the perfect length of time for the project. They then sent it to the University’s dining hall with the pellets intact, where the fresh herbs added a new flavor dimension to the students’ menu.
Next step: Data and nutrient delivery
Munir is proud that the team managed to grow a tasty crop and surpass the USDA’s national basil yields. Even though the project is officially over, he aims to scale the experiment, try other crops, and dive into the data in his final semester at Miami.
“There's a lot that can be done in terms of collecting data. Especially around nutrient delivery, using nanoparticles for fruiting crops. I'm excited to see how else I can add value.”