How the Autonomous Greenhouse Challenge Is Translating Plant Behavior Into Data
May 7, 2020 
Every grower’s ultimate goal is to maximize the yield and quality of his or her crop, while minimizing the use of resources to save costs. This is precisely the mission of the teams competing in Wageningen University’s Autonomous Greenhouse Challenge, which are doing so without ever entering the greenhouse.
So far, the challenge is progressing very well for team AuTomatoes, consisting of consultants, data scientists, engineers, researchers, and students from TU Delft, Van der Hoeven Horticultural Projects, KeyGene, and Hoogendoorn Growth Management. They harvested their first tomatoes a while ago, and the results look promising.
What is the secret of the team’s success? Who are they, and how do they do it?
KeyGene is a research company specialized in plant phenotyping and molecular breeding for the development of crop innovations. It aims to support plant breeders in optimizing crop yield and quality.
For the Autonomous Greenhouse Challenge, KeyGene contributes to the team through its knowledge of computer-based plant phenotyping and its ability to develop and exploit next-level image analysis techniques. Using cameras, plant behavior is translated into relevant data. Data scientist Niek Bouman then uses machine vision to extract the relevant data from the camera images. This data is used by Hoogendoorn to improve the artificial intelligence (AI) algorithms in order to optimize the growth climate.
Hoogendoorn Growth Management was founded more than 50 years ago, and it provides the team with experience and knowledge about horticultural automation, climate control, and data analysis. With its automation solutions, Hoogendoorn enables growers to control their growth climate in an optimal way. Hoogendoorn consultant Rene Beerkens has 20 years of experience in climate control. He shares his extensive knowledge with the team and works on building new, simple, yet intelligent controls.
Data analysts Evripidis Papadopoulos and Gerdine van Donge are engaged in the development of new and intelligent algorithms to improve irrigation and greenhouse equipment control. In addition, van Donge uses her out-of-the-box way of thinking to improve several ideas from team members, and Papadopoulos translates the feedback from the crop into data and, based on this, makes decisions about stem density and crop handling.
KeyGene also uses cameras to translate plant behavior into valuable data. This data is used by Hoogendoorn to improve the algorithms used to control the growth factors, in order to optimize the greenhouse climate according to the plants’ needs. Combined with the applied and fundamental knowledge of Van der Hoeven and TU Delft, the team possesses the multidisciplinary knowledge needed to succeed in the Autonomous Greenhouse Challenge.
You can follow team AuTomatoes on LinkedIn and Instagram to keep up with their progress in the Autonomous Greenhouse Challenge.
Subscribe to eNewsletter
