In the AGROS project, we are working towards autonomous control of greenhouse cultivation. We do so by integrating three building blocks: (1) physiological knowledge of crop growth and development processes, (2) advanced sensor technology to monitor greenhouse climate and plant status, and (3) intelligent algorithms based on mechanistic crop growth and greenhouse climate models and AI.
Greenhouse horticulture plays an important role in the year-round production of fresh and healthy products with a continuous, high quality. Worldwide, the area of protected cultivation is increasing. The production of vegetables in greenhouses must be efficient in the use of natural resources, economically viable, and produce a high quality product. This can be achieved very well in protected cultivation. However, the limiting factor is becoming the availability of sufficient highly qualified staff with knowledge of cultivation of a high-quality product and who can oversee all aspects of an efficient production system with minimal use of resources. Therefore the step to more automation in cultivation is required. In the AGROS project, we are working on the realization of an ‘autonomous greenhouse’ in which cultivation is controlled remotely via artificial intelligence, based on measurements of crop properties with the help of intelligent sensors.
First building block
The first “building block” is the physiological knowledge of crop growth and development, including underlying processes such as photosynthesis, water uptake and transpiration, and assimilate partitioning. The question we started with is which plant traits are important and should therefore be determined to be able to use them in the autonomous control of the greenhouse. An extensive list of plant traits was made, and the four most relevant ones were selected to serve as input for the intelligent algorithms.
The second building block
The second “building block” are the sensors to determine greenhouse climate and plant traits. In AGROS, we started using a wide range of sensors, ranging from temperature and humidity sensors, to sensors measuring water uptake and plant growth to vision technology for detection of plant traits such as leaf area and number of newly formed leaves. In a number of greenhouse cucumber trials, these sensors are evaluated, and in the coming period the vision technology will be further developed.
The third building block
The third “building block” are the intelligent algorithms. In the AGROS project, robust and scalable automated control algorithms will be developed that are able to generate decisions for both greenhouse climate and crop management. We will select intelligent algorithms, train these in a virtual environment based on the mechanistic crop and climate model KASPRO – INTKAM of the business unit Greenhouse Horticulture of Wageningen University & Research, and we will validate the algorithms in greenhouse trials.