'Only algorithms can handle all crop data'

Poul Erik, can you describe what you mean by autonomous growing?
"When you read about it, often people talk about data driven growing, but it's not the same. We have had a lot of data in greenhouses for many years. Maybe we do not always use it well enough. Autonomous growing is much more than data driven growing. It's a new way of using the data we already have. It's too much for a grower to work with all the data all the time. After 30 years of crop experience, I know how complex it can be. Nowadays, algorithms can do this work for them. With autonomous growing we are getting closer to the crop. Instead of yearly, weekly  or daily planning, you can go for a planning from hour to hour or even faster. We don't want to look back on failures of yesterday, we need to learn from mistakes, not to make them again. Autonomous growing supports that."

What advantages do you see in working with algorithms in your daily job?
"In general greenhouses are getting bigger and there are less experienced growers in the near future. And as a crop advisor I cannot be everywhere all the time. The algorithms are a guideline to support growers, so I can always be on top of the situation, no matter where I am. My visits to the customer can then be even more effective. Algorithms support growers in their daily decisions. As advisor I can focus on the crop strategy of the grower instead of adjusting the setpoints in the climate computer. As a grower or an advisor, you need to have the strategy clear. This depends of course on the type of crop and which variety is grown. When the right strategy is in place, the algorithms help you reach the targets autonomously in terms of desired production and quality."

Where and how do you think autonomous growing will be applied in horticulture? What still needs to be developed?
"Autonomous growing is not for small scale growers yet. It's mainly for bigger growers with more compartments or for new projects or investors who want to have a guaranteed production. They use it to help to manage the crop in bigger areas or multiple  locations. What still can be developed is including all kinds of plant measurements or sensor data.

It is not always an even situation in the greenhouse for all plants, we try to make it even, but temperatures can vary and so can crop-work, and the pressure from pests and diseases. This needs to be taken into consideration. How can the algorithms handle this?"

What's your advice to growers and investors who may still have doubts about growing with algorithms?
"To get experience in autonomous growing...you just have to start. Of course, it's important that growers trust the algorithms and the system. For growing trust, they need to learn and to get aware of what the system can do or cannot do for them. The more the grower trusts the outcome of the advice or recommendations of the models the more they can let go of setting their own setpoints."

What tips can you give Blue Radix to develop algorithms further?
"As a crop advisor and former grower, I have a lot of greenhouse experience. But for the future it's not only about experience, but also about plant knowledge and data usage. It's very important to know how data should be used and for what purpose. It's all about setting priorities in your crop.
My tips for Blue Radix are:

ˇ     Grower and advisor should give priorities how to control the settings. For example, how to reach a certain fruit quality or production target with the right fruit size. How to keep balance in fruit load and RTR. I hope that companies like Blue Radix can help us to make these decisions faster.  

ˇ     Integrating the use of sensors which can give more insights in plant status and use of smart camera's which can make predictions on yield and plant development.

The art of growing is all about details. We need to make better decisions to get better results. Algorithms can help us with that. We have so much available data now and even more in the future, only algorithms can handle this."

Background Poul Erik Lund
I started in tomato production 30 years ago in Denmark as a crop manager on a big nursery with pot plants and 2 ha of tomatoes. That was a good learning time. In the early 2000s I came to Alfred Pedersen and started as crop advisor in 2013 at Grotek Consulting. In 2015 I became co- owner of Grotek Consulting. My working area is mainly Scandinavia, Baltics and Poland. The majority of my time I spend in Poland, and Norway, but also  Denmark and Sweden is important for the company. I mainly advise tomato and cucumber growers with focus on tomato (lighted and unlighted).

Source: Goedemorgen