Review article examines AI’s potential to strengthen controlled environment agriculture

FRANKFORT, Ky. — The future of farming may depend as much on data as it does on soil, water, and sunlight.

A new review article co-authored by Kentucky State University researchers examines how artificial intelligence can help producers grow crops more efficiently in controlled environments, including greenhouses, high tunnels, vertical farms, hydroponic systems, and other indoor production settings.

The article, “Use of Artificial Intelligence in Controlled Environment Agriculture: A Review,” was published in the Journal of the American Society for Horticultural Science. The work explores how artificial intelligence can support controlled environment agriculture by analyzing large volumes of data from sensors, imaging systems, automated equipment, and crop-monitoring tools.

Kentucky State contributors include Alex Kofi, Dr. Theoneste Nzaramyimana, Dr. Adekunle Adeyeye, and Dr. Hattie Makumbe, all of the Department of Agriculture and Natural Resources in the College of Agriculture, Health, and Natural Resources. The review also includes David Kopsell of Illinois State University and James Altland of the U.S. Department of Agriculture Agricultural Research Service.

“Artificial intelligence is not replacing the farmer or the researcher,” Kofi said. “It is giving us another tool to understand crops, environments, and production systems more clearly. As first author, I wanted this review to help connect the science with practical questions growers and researchers are asking — how we can detect problems earlier, use resources more efficiently, and strengthen food production systems for the future.”

Controlled environment agriculture allows growers to manage key production factors, including light, temperature, humidity, carbon dioxide, nutrients, water, and growing space. When paired with artificial intelligence, those systems can become more responsive, using data to help forecast crop growth, detect disease, optimize irrigation and fertigation, manage lighting, support automation, and reduce waste.

The article notes that AI-related tools such as machine learning, image processing, robotics, deep learning, Internet of Things technologies, and neural networks are increasingly being used to address agricultural challenges. Those tools can help monitor crop health, soil and water conditions, climate factors, resource use, weed pressure, harvest timing, and market forecasting.

For Kentucky State, the publication aligns with the University’s growing emphasis on applied research, food systems innovation, and workforce preparation in agriculture, STEM+H, and emerging technologies. As Kentucky’s 1890 land-grant university, Kentucky State continues to advance research that connects scientific discovery with practical needs facing producers, communities, and the agricultural economy.

The review also identifies challenges that must be addressed before AI can be fully integrated across the agriculture sector. These include the need for reliable datasets, refined models, technical training, data privacy protections, and clear frameworks that give farmers appropriate control over data generated from their operations.

Even with those barriers, the authors conclude that AI-supported controlled environment agriculture offers a promising path for improving productivity, sustainability, and resilience as producers face land constraints, climate pressures, labor shortages, and growing demand for food.

The work was supported by an Evan Allen Grant: Project number 7004987 and the USDA Grants for Pioneering Controlled Environment Agriculture Research.