Blog | Data Expo

Rijk Zwaan: From better data to better cucumbers

Written by Data Expo | Sep 19, 2024 10:53:35 AM

With 3,900 employees in 31 countries, Rijk Zwaan is celebrating its 100th anniversary this year as one of the largest global players in its sector. Founder Rijk Zwaan opened a store selling vegetable and crop seeds in Rotterdam in 1924. Later, he began developing new varieties, a tradition carried on by subsequent generations. To this day, the family business continues to develop seeds for vegetables and fruits, which are sold worldwide. The company’s annual revenue is around 600 million euros, with 30% of those earnings invested in research and development each year, according to Rick van Zijl, program manager responsible for data management and data governance at Rijk Zwaan.

Challenges and Supply Chain Innovation

Continuous development is more essential than ever," Van Zijl stated at the start of his presentation, highlighting the numerous challenges in the sector: the growth of the global population, malnutrition, climate change, pressure on biodiversity, and rapidly changing consumer needs. "In our production, we take all these trends into account. We innovate for all the players downstream in the supply chain, starting with us: breeders, processors, retailers, and consumers. We focus not only on how well crops can be produced but also on how effectively they can be packaged. Further along the chain, it’s about aspects like shelf life and flavor.’

More data, more objective data

The demand for innovative and resilient crops is high in 2024. A complicating factor: it takes seven to fourteen years to breed a new variety, such as a larger tomato, a differently colored pepper, or a new type of snack vegetable. During the development process, a wealth of data is collected, from plant characteristics to growing conditions, growth indicators, and DNA data. Van Zijl: "We want to continuously expand the amount of data to make better decisions." However, he added, it's not just about quantity: "Besides more data, we especially want more objective data. For example, you want to avoid a cucumber's quality being judged differently on a rainy Monday morning than on a sunny Friday afternoon."

Digital Phenotyping

This brought Van Zijl to phenotyping, the process of mapping an organism's external characteristics. Rijk Zwaan has fully digitized this: manual field assessments have been replaced by smart observations of crops and their traits. For example, Rijk Zwaan takes advanced digital photos of vegetables on a (mobile) conveyor belt to measure length, weight, color, and other attributes. The company also uses drones to scan fields and monitor growth and health. In addition to cameras, algorithms are used to capture and analyze data on the physical and biochemical characteristics. Van Zijl: "It all generates useful data that we aim to use as scalable and user-friendly as possible." A robust data platform was essential to achieve these ambitions, which led Van Zijl to HSO, a company Rijk Zwaan already collaborated with.

Data-driven decision-making

For HSO, phenotyping was a new concept, admitted Jerrold Stolk of the Microsoft service provider at the Data Expo: "To assist Rijk Zwaan effectively, we first needed to understand what phenotyping entails and assess the requirements for digital phenotyping at scale," said the Tech Lead Data & AI. "We mapped all data points worldwide, from Guatemala to South Africa, and where necessary, created cloud-to-cloud integrations. Then we standardized the data. By measuring more accurately and efficiently, we can optimize selection and help breed superior plant varieties." HSO developed a structured step-by-step process for Rijk Zwaan, turning raw information from photos and metadata into decision-making tools, answering questions like: what makes a good cucumber?

Azure-cloud

Van Zijl went on to explain how Rijk Zwaan, with HSO's guidance, began leveraging Azure Cloud, where the Microsoft partner has extensive expertise. It facilitates scalability and self-learning capabilities, preparing the seed breeding company for the next hundred years. Van Zijl concluded: "We always want to provide our growers with the best possible information. Machine learning helps us grow the best vegetables and fruits together."