Greenhouse growers have a number of tools available for collecting data through all stages of plant production. For many, the next step is understanding how that data can be used to build strategies for improving operational efficiencies.
Ken Tran
During the GROW Executive Summit, Ken Tran, the founder and CEO of Koidra Inc. and formerly a Principal Applied Scientist in the Machine Learning Group at Microsoft Research, offered a glimpse at a potential future of autonomous greenhouse production, with all decisions being made based on in-depth data analysis.
If you’re a grower and business owner today, you might be wondering how information like this can be used to solve today’s problems. According to Tran, it starts with labor, although perhaps not in the traditional sense.
“Everyone talks about labor savings, but the kind of labor we are trying to address is your highly skilled and highly technical talent at the engineering level,” Tran said. “This kind of labor is very hard to find, and even the top growers still have room for improvement.”
This lack of technical talent is compounded when growers want to make more data-driven decisions, but the data they are collecting (if they are even collecting it) is fragmented or stored in programs that may not communicate with each other.
“It’s a data hostage issue,” Tran says.
At the highest level, Tran says, companies are successful or not successful because of the decisions that are made at different levels of production, and the logic being used to make these decisions.
“Operational decisions, either in real time or near-real time, start with your team of experts,” Tran says. “That’s great, but if you can influence the decision-making process inside a greenhouse with the addition of sophisticated models that can distill actionable insights from the data you collect, that’s what artificial intelligence (AI) is all about.”
In order to leverage AI effectively, Tran says, you need to be able to automate both the data collection and the decision making.
“It enables you to combine AI with the Internet of Things, and every piece of technology used in the greenhouse, from cameras to sensors, can be useful,” Tran says.
The market for this type of technology is going to accelerate rapidly over the next five to ten years, Tran says. Because of that, business owners need to be the ones making the decisions on which technology to use.
For a deeper dive into the layers of digital horticulture, be sure to check out Tran’s article on GreenhouseGrower.com
