Does Using More Smart Sensors Also Lead To Better Cultivation Results?

In a few months, it will be clear which team wins the Autonomous Greenhouse Challenge and has therefore developed the best AI algorithm for autonomously controlled cultivation of dwarf tomatoes. Parallel to the participants, a team of WUR experts has also set up a greenhouse for autonomously controlled cultivation. Researcher Pinglin Zhang explains the use of this reference greenhouse and why the extra sensors and cameras that the teams use can also provide valuable knowledge for WUR.

In Bleiswijk, dwarf tomatoes are currently growing in five greenhouse compartments that are controlled by AI algorithms, developed by the teams of the Autonomous Greenhouse Challenge . Dwarf tomatoes are also growing in a sixth compartment, but then controlled with input from a team of WUR experts. One of these is researcher Pinglin Zhang. “We also call our compartment the reference greenhouse. The cultivation is also controlled autonomously here, but without AI algorithms. We have set up the greenhouse in the same way that a regular grower would control the greenhouse. With this reference, we can make a good comparison between the results of greenhouses that are controlled with AI algorithms and a greenhouse that is controlled in a conventional autonomous way.”

Additional sensors and cameras
Another difference is that, unlike the WUR team, the participants have been allowed to use additional sensors and cameras since this year, says Zhang. “Basic sensors are installed in all six compartments – including ours – for measuring the temperature and relative humidity in the greenhouse and real-time weather conditions outside, for example. Some teams have installed their own sensors and cameras on top of this. Think of a sensor that measures the weight of a pot. Or a thermal camera that measures the temperature around the leaves and fruits. The teams therefore have more input that they can use in the control than we do in our reference greenhouse.”

Data processed by algorithm
The teams have developed their algorithm in such a way that the data collected by the sensors and cameras are processed directly to make an optimal decision. Zhang: “A low weight can indicate insufficient water, which means the plant can be given extra water. Information about the weight can also be used to determine the harvest moment. Just like the redness of the fruits that the cameras register. Based on the strategy of the teams, the algorithm calculates the harvest date. Whether this ultimately leads to the most productive harvest with the best quality fruits will become clear at the end of the challenge.”

From greenhouse level to plant level
As a researcher, Zhang focuses on greenhouse technology for more efficient and sustainable crop cultivation. “Data from sensors can help growers optimize energy consumption for lighting, heating and ventilation. In addition to the greenhouse level, I also look at sensors on a smaller scale. For example, I am working on a project measuring the microclimate: the temperature and humidity around a plant. By placing sensors close to the plant, you collect data about the local conditions. This can give an indication of, for example, the presence of bacteria that can cause diseases. If you have that in mind, you can intervene at an early stage.”

More attention for the plant
Zhang also sees increasing attention for monitoring the plant with smart sensors and cameras in the Autonomous Greenhouse Challenge . “In the past, the focus was almost exclusively on the greenhouse climate. That provides valuable data for controlling the greenhouse, but gives less information about how the plant itself grows. While that is ultimately the most important thing for good production. Some teams really work with state-of-the-art technology . For us at WUR, it is also very interesting and valuable to see the latest developments in the field of sensors and cameras.”

Future of autonomous cultivation
Finally, how does Zhang see the future of autonomous cultivation? “That depends on how you look at it. There are different degrees of autonomous control of a greenhouse. For example, autonomous regulation of the climate is already well embedded in greenhouses in the Netherlands and other Western countries. If you are talking about autonomously determining cultivation strategies or applying robotic techniques to crop management, then that is still in its infancy. That is one of the things we hope to explore further with the Challenge. How far we have come exactly and whether fully autonomous cultivation is even possible, we actually do not know yet. That is also what makes this very interesting.”