Sjoerd Raben, Project Manager, Software Development, Royal Kaak (pictured: right), offers insights into
the company’s work, perspectives on AI concepts, and advances in incorporating game-changing tools in the solutions provided.
BBI: What are the highlights of your R&D work in AI over the past three years?
Sjoerd Raben: Over the past three years, our R&D efforts have focused on integrating AI capabilities into a scalable cloud-based data platform for industrial bakery environments. A key achievement has been enabling structured data collection across equipment and production lines, creating a solid foundation for advanced analytics.
This has allowed us to move from isolated machine data towards a connected ecosystem where performance, quality, and efficiency can be analyzed holistically. By combining industrial IoT with data-driven insights, we support bakeries in improving consistency, reducing downtime, and making better-informed operational decisions.
BBI: In what production areas have you already incorporated AI features into your technology?
Sjoerd Raben: AI is primarily applied in the area of data-driven decision support, where we use generative AI to help users interpret complex production data and translate it into meaningful insights.
Instead of presenting raw data or dashboards alone, generative AI enables operators and production managers to better understand what is happening in their process, why certain deviations occur, and where improvement opportunities lie. This significantly lowers the barrier to using data effectively on the shop floor.
In addition, we support predictive maintenance through advanced data analysis, where patterns in equipment behavior are used to anticipate potential issues. This is based on structured data analysis and domain knowledge rather than autonomous AI-driven decision-making.
“Importantly, control of the production process remains with the operator. This approach strengthens the sense of control on the shop floor, as decisions are supported by transparent insights rather than relying on a ‘black box’ system that autonomously adjusts process parameters.”
Sjoerd Raben, Project Manager, Software Development, Royal Kaak
BBI: What are the most frequent requests you are receiving?
Sjoerd Raben: The most frequent requests focus on improving operational efficiency, ensuring consistent product quality, and gaining better visibility into production processes.
Customers are increasingly looking for ways to make their processes more transparent. They want to understand how different parameters influence product quality and how to react more quickly to deviations.
There is also a strong demand for solutions that simplify complexity – tools that make data accessible and understandable for operators, rather than adding additional layers of technical interpretation.
BBI: How is machine learning defined to optimize various production steps in real time?
Sjoerd Raben: At this stage, machine learning is primarily used for analytical purposes, rather than directly controlling production processes in real time. By analyzing historical and real-time production data, machine learning helps identify patterns, correlations, and deviations that are not easily visible through traditional methods. These insights provide operators with a clearer understanding of their processes and highlight areas for improvement.
Importantly, control of the production process remains with the operator. This approach strengthens the sense of control on the shop floor, as decisions are supported by transparent insights rather than relying on a ‘black box’ system that autonomously adjusts process parameters.
Data quality is essential
BBI: What are the challenges with AI decision-making algorithms and how are consistent results ensured?
Sjoerd Raben: One of the key challenges is ensuring reliability in complex and variable production environments. Bakery processes are influenced by many factors, including raw materials, environmental conditions, and operational variations.
To ensure consistent results, we place strong emphasis on data quality, robust data pipelines, and clear system monitoring. AI is used to generate insights based on this data, but always within a controlled and well-understood framework.
Another important aspect is transparency. Insights must be traceable and understandable for operators, so that they can confidently act upon them and maintain full control over the production process.
BBI: What opportunities do you see in expanding AI use in bakeries?
Sjoerd Raben: We see significant opportunities in expanding AI within bakeries, particularly in making production processes more transparent, efficient, and consistent.
At the same time, we believe that successful adoption requires a step-by-step approach, where AI is applied in small-scale, clearly defined, and comprehensible use cases. This ensures that human control over the overall process is maintained at all times.
A critical prerequisite for this development is the availability of clean, structured, and well-labeled data. In many cases, simply organizing and utilizing this data – without applying AI – already provides substantial value. AI should build on this foundation, not replace it.
“We believe that successful adoption requires a step-by-step approach, where AI is applied in small-scale, clearly defined, and comprehensible use cases. This ensures that human control over the overall process is maintained at all times.”
Sjoerd Raben, Project Manager, Software Development, Royal Kaak
BBI: What does your R&D prioritize in further AI advancements? And how are quality and safety concerns covered?
Sjoerd Raben: Our R&D priorities focus on scalable data infrastructure, practical usability, and reliable insights. We continue to invest in strengthening our data platform to ensure that it can support growing volumes of production data while remaining accessible and useful for operators and engineers. At the same time, we focus on developing solutions that are intuitive and fit naturally into existing workflows.
Quality and safety remain central. AI is applied as a supporting layer that enhances understanding and decision-making, while established control systems and human expertise remain responsible for final process control. This ensures that innovation goes hand in hand with reliability and product safety.