Managing natural resources in greenhouse production is a top-priority issue. Water- and nutrient resources are becoming increasingly scarce and expensive. In order to further optimise their resource efficiency, we are working on developing plant sensors which will enable us to adapt the quantity of resources supplied to the crops according to their actual needs.
The ‘networked plant’ concept is based on the continuous real-time measurement of the electrical signals flowing through the plants. The raw data collected are evaluated with artificial-intelligence-based – or more generally with machine-learning based – methods enabling the needs of the plant to be modelled.
So far, the water requirements of a hydroponic greenhouse tomato crop have been modelled, and an automatic irrigation controller is currently being tested. Parallel to this, work is being done on early mineral-deficiency detection. An application that aims to detect ‘general stress’ has been developed as a practical tool for greenhouse production.
This new method is also being refined for use in other crops such as cucumbers, aubergines, basil, grapevines and apple trees.