What if there are two greenhouse compartments, both equipped with a commercial greenhouse control system to maintain climate conditions, both growing cherry tomatoes under identical planting densities and maintenance protocol – but one is managed by AI and the other one by an expert grower? That was the goal of a trial that spanned from the end of June to mid-November at the Ohio State University's Controlled Environment Agriculture Research Complex (CEARC). The results? The AI system of Koidra outperformed conventional methods with a 9.6% increase in total yield over 13 weeks of production.
© Koidra
The results were presented as one of 10 presentations selected for a plenary technical session at GreenSys 2025, a leading international conference on greenhouse innovation.
"Greenhouse management is an intricate dance of balancing temperature, light, humidity, and nutrients - each factor influencing yield, resource use, and crop quality. Conventional methods rely heavily on growers' expertise, demanding constant vigilance and adjustment", the Koidra managers explain. "Yet, even the most skilled growers face challenges, such as rising energy and resource costs, straining operational budgets. Inconsistent environmental conditions lead to yield variability, but data-driven decision making remains elusive without advanced tools."
Funded by USDA SCRI grants, The Ohio State University collaborated with Koidra to address these pain points, implementing an AI-driven climate control system that promised to not only optimize conditions but also support growers in making data-informed decisions.
Koidra's KoPilot solution leverages physics-informed machine learning (PIML) to refine greenhouse management to adjust lighting, temperature, humidity, and CO2 levels for optimal photosynthesis and transpiration. The system integrates real-time data and predictive algorithms to create a dynamic, responsive environment tailored to crop needs. "KoPilot's technology is developed to enhance growers' decision-making on top of your growing expertise", the team says.
© Koidra
This scientific-grade comparative study was designed to assess the practical impact of autonomous climate control on cherry tomato production. It is the first known of its kind in North America and only the second globally in the context of greenhouse cherry tomatoes. The trial spanned from the end of June to mid-November, comparing an AI-controlled greenhouse compartment with an expert grower-managed one. The two 482 m2 compartments were equipped with a commercial greenhouse control system to maintain climate conditions. Cherry tomatoes are grown under identical planting densities and maintenance protocols.
"The positive outcomes showcased the potential of AI in CEA", the team shares. "AI outperformed conventional methods with a 9.6% increase in total yield over 13 weeks of production."
As insights for expert growers, the power of precision was addressed. "By continuously analyzing environmental data, AI-driven systems fine-tune conditions in ways that maximize yield and quality." This innovation opens sustainability doors. "As AI technology evolves, resource use is further optimized to align with sustainability goals." It's about partnering with technology, is the conclusion. "This trial demonstrated that AI empowers growers. By automating routine tasks and providing actionable insights, growers can focus on strategic planning & decision-making."
For more information:
Koidra
[email protected]
www.koidra.ai
