For mushroom growers, keeping track of crop development is an ongoing challenge. Mushrooms do not grow uniformly, and unlike many other crops, their status is difficult to observe directly. Belgian company Heliovision, based in Leuven, developed the AI application Funghii to support growers through data measurement and harvest prediction. "The market need was clearly there, as shown by how quickly Funghii has been adopted," says Emile Derache of Heliovision.
© Heliovision
Heliovision is a software company specialising in machine vision solutions based on advanced 2D and 3D imaging. Founded five years ago, the company develops custom software aimed at advancing automation. "We programme cameras so they can automatically detect and analyse objects," Derache explains. "This is primarily used in quality control and measurement applications within industrial processes."
Challenging growing environment
The company operates across several sectors, but decided to focus specifically on the mushroom industry with Funghii. "Growers told us that mushroom cultivation was becoming increasingly precise and complex, and that it was getting harder to monitor and predict everything accurately," says Derache. "During earlier projects, we were repeatedly asked by mushroom growers to install cameras in their growing cells. From a technical standpoint, however, that is an extremely challenging environment, with high humidity, significant temperature fluctuations, and regular water use. In the early stages, we often said it was too complex to make work reliably."
© Heliovision
"But the requests kept coming," he continues. "At a certain point, you start rethinking your position. If so many different growers are asking for the same thing, there must be a solution. That's what led us to pursue it, and it has become a clear success in a relatively short time. Growers in Belgium, the Netherlands, Poland, the United Kingdom, and Germany are now working with our system."
Better insight into growth and commercial planning
Funghii works with four cameras per growing cell. Depending on the size of the operation, a typical facility has 12 or 24 cells. The cameras are positioned to monitor multiple measurement points throughout the full growing cycle of approximately 30 days. "Based on that data, we carry out a series of measurements to give growers better insight into crop development," Derache explains. "We measure growth rate and cap diameter, and we count the number of mushrooms present, which is essential for estimating what can be harvested the following day."
"Because mushrooms do not all grow at the same time, but come in distinct growth flushes, we can use that data to predict when each flush will be ready for harvest. This gives growers a much clearer picture of what will happen over the coming days, which is critical for planning and logistics."
© Heliovision
Beyond crop management, Funghii also offers commercial benefits. "It helps growers predict how much they will harvest and when. That matters a great deal in communication with buyers such as supermarket chains. If you only find out at the moment of harvesting, it is already too late. That can lead to logistical problems, last-minute calls, or even financial penalties when delivery agreements are not met. On the other hand, a surplus is also problematic, since mushrooms have a shelf life of only a few days under refrigerated conditions."
The timing of technological development
The speed at which Funghii has been embraced suggests the challenges it addresses have existed for some time. Two factors explain why a solution did not emerge sooner. "First, there was the technical difficulty of developing cameras that function reliably in such extreme conditions," says Derache. "The combination of high humidity, temperature variation, and water use is not straightforward, and many companies simply did not attempt it."
"Second, there has been rapid progress in AI, particularly in computer vision, over the past few years. This allows for far more precise and detailed analysis of what happens inside growing cells. Where that was not yet feasible a few years ago, it is now possible to detect mushrooms as small as one millimetre and track them from the earliest growth stage. The combination of robust hardware and recent advances in AI has brought everything together at the right moment."
© Heliovision
Tomatoes and courgettes
Derache points to the lack of prior automation in the mushroom sector as part of the reason Funghii has generated such enthusiasm. "In many other crops, automation has advanced considerably, but in mushrooms, that has not been the case for a long time. It is a relatively small sector, which made it commercially unattractive for many companies to develop solutions. That is exactly why there is now such strong interest when technology arrives that genuinely makes a difference in growers' daily work."
Heliovision is also exploring applications in other sectors. "We are looking at greenhouse horticulture, for example, tomatoes and courgettes. These are year-round greenhouse crops, which makes them comparable to mushrooms in several ways," says Derache. "We are also thinking more broadly, about applications in livestock farming, such as poultry and pig production. The hardware we have developed is built to withstand demanding conditions, which is equally important in livestock housing and greenhouses. That means the foundation we have built is broadly applicable, and we are actively exploring the possibilities."
For more information:
Emile Derache
Heliovision
Endepoelstraat 3A / unit 12,
3210 Lubbeek, Belgium
Tel.: +32 16 797 960
[email protected]
www.heliovision.be
www.funghii.com
