What is AutoCEA?
It is a crop-state-driven autonomous greenhouse control framework linking visual crop perception, phenological decision-making, and equipment execution.
Why this work?
Most automated greenhouses execute environmental setpoints, while decisions such as crop-stage interpretation, plant respacing, resource allocation, and harvest timing still depend on human growers. AutoCEA treats crop state as the basis for control decisions.
Approach
The framework uses a perception-decision-execution architecture. RGB observations and sensor data estimate canopy coverage, flower and fruit counts, red-fruit ratio, growing degree days, daily light integral, and drainage feedback.
These signals inform phenological staging, plant respacing, and harvest timing; the execution layer adjusts lighting, temperature, irrigation, CO₂, ventilation, blackout screens, and energy screens.
Key evidence
0.34EUR m⁻² d⁻¹ net profit in the Challenge system, versus 0.13 for the expert benchmark
−7.7%Total device cost relative to common practice derived from crop guidance and grower rules
2024Implemented in the 4th Autonomous Greenhouse Challenge for dwarf cherry tomato production
Questions this work answers
What makes it different from setpoint control?
Its control decisions are organized around crop development and production management, not only fixed environmental targets.