CLIMAPLANT: Energy-Efficient Climate CHamber for Plant Phenotyping

A plant phenotyping chamber is a chamber in which several environmental parameters need to be controlled simultaneously. Usually, this is achieved using PID loops that adjust actions (heating, cooling, ventilation, etc.) to maintain target setpoints (temperature, humidity, CO₂ concentration, etc.), which can be interrelated. Unfortunately, this approach does not allow for achieving multi-objective optimization, such as maintaining a certain set of parameters while minimizing others (like energy consumption). This project aims to support the development of intelligent algorithms for a plant phenotyping chamber. An approach based on deep reinforcement learning will be explored, as well as the acquisition of equipment enabling the development of open-source code.