The mushroom market is booming. The worldwide market size has grown 8% year over year, and yield is expected to be doubled in 2023 compared to last year (De Ligt, 2018). As more and more mushrooms are consumed, suppliers will have to scale up in order to meet this increase in demand. While this scaling occurs, maintaining the quality of the mushrooms gets harder. To help these growers, Delphy, Geurts Champignons and Widget Brain started working on automated mushroom cultivation with self-learning technologies to maximise the yield while securing quality. We interviewed two experts at Delphy, including Niek de Ligt (consultant/trainer, composting & growing) and Jan Gielen (manager/specialist climate & energy at Delphy) and asked them to share their vision and thoughts behind this development.
When the market is scaling to all-time heights, providing personal consultancy and hands-on training, as Delphy traditionally does, has its limits. To be able to scale this knowledge sharing, Delphy decided to partner up with Geurts Champignons and Widget Brain to realise supervised learning. In this case, Delphy captures the knowledge about the ideal climate conditions of growing rooms for each growing phase in their rule repository. All of these growing rooms need a tailored composition of the right climate conditions, depending on the growing phase, to optimise mushroom yield and quality. Widget Brain’s algorithms can optimise the conditions of each growing room based on other related inputs, like compost, casing, mushroom strain etc, in order to maximise the yield and quality.
Gielen recalls that the idea to digitise the growing room process and thereby optimise mushroom yield and quality has been around for a long time. In the past, he’s made plans and drafts for this concept, but it wasn’t able to be implemented until Widget Brain came into the picture: “It was only until recently that the technology behind such systems matured. It became cheaper and more accessible.”
“Being able to steer along from a distance is great, since we don’t have to be physically present at the farms that often to give consultancy or train the growers.” says De Ligt. The combination of the digital growing assistant QMS (Quality Measurement System), and support from a distance by an experienced consultant, leads to maximum efficiency and transfer of knowledge in the field of consultancy and training for mushroom growing.
Implementing this technology, however, hasn’t always been a smooth ride. De Ligt mentions the obstacles along the way, such as communication and financing between all parties. These parties include Delphy as the supplier of climate and growing knowledge, Widget Brain as the technology supplier, and Geurts Champignons as the mushroom grower. De Ligt: “We have specifically chosen to work on this project together with an actual and innovative mushroom grower. This way we work with real data and thus show the real results of the project. It should prove that it doesn’t only work in theory, but also in practice.”
Finding the right partner for this project wasn’t necessarily hard for Delphy. According to Gielen and De Ligt, many mushroom growers already knew that something had to be done with all the data they have. It was just very important to communicate the overall process, requirements from the growers and impact of the project on the grower’s profit margins. Gielen elaborated on that: “We knew we had to propose to start on a small part of the growing assistant QMS, just optimising on one parameter (evaporation). This became clear after an inventory we previously did under growers, consultants and suppliers of compost, casing and climate computers. I believe that growers gradually have to get more familiar with it. All that data, all parameters and optimising everything would have been too overwhelming.”
“Geurts Champignons also believed in the value of making smart links between their climate computers and our climate and growing knowledge. When we explained what the project would entail, they enthusiastically jumped in. That’s very important.” De Ligt says. “We know that there will be setbacks, so we need 100% commitment of all parties to make this project a success.”
“All the knowledge we gained in the last 50 years to facilitate the market, has to be positioned abroad in the coming 5 years to support the double growth. That’s 10 times faster! We need the technology and this system to scale accordingly.” says De Ligt. One way to scale in such manner is by planning it wisely. Gielen: “Another reason why we chose Geurts Champignons as project partner was that they are using a brand of climate computers that is commonly used all over the world. If this project works on this brand of climate computer, we scale much easier and later implement it with other brands of climate computers.”
Gielen and De Ligt elaborate further on the future of their business processes. Currently, Delphy mainly provides training and consultancy in a traditional sense, where they are physically present at the locations to provide guidance on composting and growing. De Ligt: “This will be expanded with the QMS system, which we call the digital growing assistant, which will provide a part of the knowledge transfer digitally. The trainees will go through a shorter learning period with a steeper learning curve. The traditional on-site training and consultancy will stay, but we are definitely moving towards digital knowledge sharing.”
Digital knowledge sharing is exactly what they need to keep up with the demand and growth of the market. As the number of companies shrink and standing companies grow, knowledge and expertise become rare. Gielen explains how large foreign companies have hired multiple advisors in the past, simply because local experts were not available. “That’s really expensive for mushroom growing companies. We wanted to provide a better solution and now we capture all that expertise and knowledge into a system that will be widely available and doesn’t require as much human interaction.”
To go a bit deeper into the future of the business, Gielen portrays future business models. He explains how face-to-face trainings and consultancy are time consuming and sometimes expensive for growers. Therefore, growers often only start taking action after they run into a problem. That’s why Delphy usually works on a project basis. In the future, however, it might be possible to get recurring revenue from the QMS and the consultancy that goes with it. “Digital consultancy is cheaper because it’s not human-intensive. Besides that, the consultancy that rolls out of QMS is remotely monitored, controlled and scalable. A subscription-like fee would then make more sense. The human advisors will stay, but they will be more like a consultant where they know how QMS works and compliment it.”
For the immediate future, Gielen speaks with confidence about expanding the current functionalities for the QMS. “When the project works and everything goes well for one parameter, I’m sure that the growers will get very enthusiastic about expanding the QMS with more parameters because that data is already available inside the system. I’m sure of it.”
This content is owned by Widget Brain and written in collaboration with Delphy. Widget Brain is a software company specialised in algorithms, machine learning and IoT for the Industrial Equipment industry, Retail, Maritime and Waste. Learn more about Widget Brain here.
By joining adaptive algorithms with service planning, scheduling maintenance becomes automatic, efficient, and cost-effective. Relying on machine data, algorithms can help companies make the transition from a reactive to a proactive scheduling model and resolve maintenance issues before they disrupt production. Read more.
Increasing competition, rising skill shortages and high turnover rates are tormenting the hotel industry. In order to survive, hotels have to elevate hotel experiences of their guests and provide the best possible service which is enabled by staff to survive. So, how can hoteliers enable their workforce to elevate guest experience? With AI-driven scheduling.
If you want to improve your machine's performance and output quality, Performance Analytics and Optimisation is a way to do that. You will not only achieve higher KPIs like units/hour and yield (%), but also get a better understanding of how your machines work and the ability to predict future performance. Read more.