Biobest embraces AI to help growers drive up yields

30 years ago, Roland De Jonghe of Biobest pioneered the commercial production of bumblebees and their use for pollinating greenhouse tomatoes - not only setting off a revolution in crop pollination but providing an early driver for biological control and integrated pest management (IPM). Fast forward three decades and the company he founded - Biobest - is a global leader in both areas. 

A pioneering company with strong beliefs that robotization, artificial intelligence (AI) and digitalization are set to shape the future of greenhouse growth, Biobest has recently announced two new partnerships with companies leading in these fields.

Data-driven growing 

"In simple terms agriculture involves drilling seeds in a field, adding inputs - such as fertilizer, agrochemicals and irrigation - and then largely leaving mother nature to get on with it," says Karel Bolckmans - COO of Biobest. "In contrast, growers utilizing high-tech glasshouses have far more control over their plants - they can steer crops minute-by-minute, manipulating parameters including temperature, relative humidity, CO₂ concentration, light, electroconductivity (EC) and substrate pH.

"Data-driven growers constantly measure these parameters to keep a grip on plant growth - using it to steer vegetative or generative growth - and therefore optimize crop yields. 

"Growers base their decisions mostly on 'gut feeling' and experience and their 'reading' of plants," says Karel. "In the Netherlands, we have the Autonomous Greenhouse Competition - an annual event which sees leading growers pitted against AI to make crop management decisions. The computer always wins!"

Using data analytics to improve IPM 

Turning to IPM, crop scouting remains a vital component of biological control strategies. While highly labor intensive, it is also selective - as it is not viable to monitor every plant. 

"Our recently launched Crop-Scanner App, which we developed in close collaboration with the Israeli company Agritask, facilitates recording and visualizing scouting data - helping customers make better decisions in their crops," says Karel. "The next logical step is to improve early detection of pests and monitor their evolution in the greenhouse using sensing cameras and computer vision technology."  

With this in mind, Biobest has recently entered a new channel partnership with ecoation - the Canadian developer of automated greenhouse management and crop health monitoring technologies. As a result, Biobest is now offering ecoation's OKO machine - centered around closed-loop IPM, yield production assessments and crop work quality checks - to growers in Europe.  

Explaining how it works Karel says, "Driven around the greenhouse on the pipe-rail system, this OKO scouting robot looks at each plant. It is programmed to detect abnormalities by recognizing biotic (biological) and abiotic (environmental) plant stresses. 

"Where plants are showing sign of stress, OKO can detect abnormalities invisible to the naked eye. It uses this data to create 'heat maps' of the greenhouse. Accurately pinpointing areas where a plant, or group of plants, is showing sign of stress and underperforming - it also records the cause.

"This is a huge step forward. Until now, modern environmental-controlled glasshouses have typically had a single static sensor - recording climatic parameters. OKO is a dynamic sensor, recording precision data regarding temperature, relative humidity, CO₂ concentration, light intensity etcetera throughout the glasshouse - creating a high-resolution map for each parameter."

He goes on to explain, "All this data can be layered on top of each other to provide answers to questions such as, how fast will a particular crop pest develop and in which direction - to inform decision-making." 

Digitization of IPM 

"Ecoation technology allows for the digitization of the IPM workflow. OKO provides an essential tool for the rapid information gathering and knowledge distribution necessary for the timely optimization of biological control strategies.

"At Biobest we set out to provide 'best in class' technical support. Many customers receive a visit from one of our technical advisers at least every two weeks. While highly labor intensive, the advent of damaging plant viruses - such as Pepino Mosaic Virus (PepMV) and Tomato Brown Rugose Fruit Virus (ToBRFV) - as well as the human virus COVID-19, means regular visits are not always ideal or permitted. 

"Particularly during the last 18 months, we have seen a growing need to provide the best possible technical advice in the absence of regular visits. To draw the right conclusions remotely, we need access to powerful and accurate data harvesting, using artificial intelligence, algorithms, predictive analytics and decision support systems. 

"Supplying all of this, OKO also has the potential to provide yield predictions and can objectively assess the quality of crop work such as fruit picking. These are powerful additional tools. In short, OKO informs integrated crop management and integrated pest management decisions based on hard data."  

With OKO, growers can improve the climate in the glasshouse as well as pest and disease control using IPM. While benefiting from cost savings, by improving the IPM programmes growers spray less frequently so plants are less stressed helping optimize yields. 

"With a worldwide trend towards larger-scale growing operations, involving multi-sites, OKO can help growers deal with this complexity," says Karel. "Major growers can no longer afford to rely on their gut - they need hard data to do the best job possible."

Karel predicts data-driven IPM and precision growing will be the next big revolution for the industry. "Growers will be managing at a plant, rather than a crop, level," he says. "We are entering a new era involving the integrated use of cameras, decision-support systems and analytically tools - big changes are coming."

OKO is initially to be launched in Belgium and the Netherlands, with plans to expand into Continental and Eastern Europe. An autonomous model is due to be launched in 2022. 

Pollination robots 

Continuing on the theme of AI and digitalization, in June Biobest announced a new strategic alliance with Arugga AI Farming of Israel - which will see it take the lead in robotic pollination in North America.

According to Karel Bolckmans, there are certain situations where robotic pollination can work more effectively than bumblebees. "If the technology is better - you cannot stop it," he says. "It will always win and, rather than fight it, it is better to embrace and work with it. In high-tech tomato crops, pollination robots can significantly increase yields compared to bumblebees." 

In comparative trials in high-tech tomato crops, Arugga's pollination robots have boosted pollination levels by 2-5%, compared to bumblebees. This higher fruit yield equates to significantly higher returns per hectare. 

In many parts of the world, high-tech greenhouses are turning to artificial light and shade screens to manipulate and optimize growing conditions. However, these systems can negatively impact bumblebees causing orientation problems.  

"The bumblebees can find it difficult to navigate back to their hives," explains Karel. "This is particularly an issue in speciality tomato crops - which require two to three times as many bees compared to standard round and beef tomatoes."

During extremely hot periods, the pollination robot comes into its own. "In high temperatures, flower and pollen quality can decrease," explains Karel. "These flowers require significantly more vigorous shaking to loosen the pollen.  

"We are currently setting up four major trials with leading tomato growers in North America. We are confident the results will be good. There is no doubt, pollination robots have a bright future in high-tech tomato crops. From a practical standpoint, the robot needs a parallel pipe-rail system to run along and, of course, the economics must add up. There are plenty of low-tech greenhouse tomato crops in the Mediterranean and elsewhere where bumblebees will continue to play a vital role in pollination and ultimately yield."  

No-touch pollination 

Travelling between rows, the Arugga Pollination Robot's cameras recognize each tomato flower cluster and points an air gun at it. This shoots a calibrated pulsating air current at the cluster - vibrating the flowers and loosening the pollen.  

"The robot mimics the buzz pollination of bees but, importantly, at no time does it touch the plant," says Karel. "Plant viruses - such as PepMV and ToBRFV - can be mechanically transmitted within a crop so 'no contact pollination' is ideal. Indeed, the robot is coated with a special disinfectant paint.

"Looking to the future - while the robot is moving around the crop it has the potential to carry out other functions, to multitask."  

Future for bumblebees 

As to the future of bumblebees, Karel says they will still play the key role in countries reliant on less sophisticated glasshouses technologies - as well as for pollinating crops such as blueberries, melons and strawberries, which are not reliant on buzz pollination. 

"At Biobest we set out to provide growers with the very best solutions. Before Biobest pioneered the commercial production of bumblebee hives, growers of protected crops were largely reliant on manual pollination. Highly labor intensive, they had to vibrate plant clusters individually using hand-held devices, tap crop wires or resort to spraying synthetic plant hormones to stimulate fruit set - producing fruits without seeds. 

"Biobest bumblebees revolutionized pollination back in the 1990s. Thirty years on and we are still as enamoured with our bees! While bumblebees are set to continue to play a vital role pollinating many high-value crops, we are thrilled to be a part of the next revolution in tomato pollination using robots."

Source and Photo Courtesy of Ag News

Source: Ag News