The SmartMill journey – one smart step at a time  │Inspiration Hub │Bühler Group

Artificial intelligence (AI) has not suddenly emerged out of nowhere. Techniques such as machine learning and data‑driven optimization have been part of industrial solutions for years, including at Bühler. What has changed in recent years is accessibility. As Alexis Noel, Head of Product Management, Smart Factory at Bühler, puts it: “We’ve been working with data for decades – collecting it, recording it, integrating it. What is new is how easily we can now combine data to look at the bigger picture.” 

Instead of focusing only on individual machines or parameters, digitalization makes it possible to analyze entire processes and operations across multiple sites. “We’re no longer just looking locally at one machine or one process,” says Noel. “We are making sense of the overall business.” 

This shift is the foundation on which SmartMill is built – from connecting to monitoring, assisting to self-optimizing. 

A well-planned journey 

“The SmartMill is a journey of transformation, not a disruption,” says Georg Schafler, Head of Wheat and Rye Process Technology at Bühler. “Our customers are right to be cautious about technological promises. We know that in milling, investments are long term, margins are often tight, and reliability is critical. That’s why we support customers at each step of the journey and focus on the returns on investment.” 

Schafler works closely with customers and operators embarking on this journey and says that a gradual approach is essential. “We tackle the topic step by step. Not everything at once,” he says. 

The first steps focus on connecting existing equipment and making data visible. With a digital record of production, dashboards, and analytics, the system provides actional insights for improving yield, quality, and energy use. These first steps are intentionally low risk. By using data already available from the plant control system, mills can gain transparency without major additional investment. Operators begin to see patterns, deviations, and trends – and gradually become comfortable working with digital tools as part of their daily routine. 

From this foundation, SmartMill builds toward more advanced capabilities. The next step is to move toward self-optimization for a selected function. Once this is established, the whole system can become self-optimizing: with the help of continuous learning, the system adapts parameters to the specific characteristics of the raw materials and the conditions in real time. 

The difference that AI makes is that it can go beyond a single task with a single solution and can take over many tasks in real-life situations with complex and changing parameters. For millers, dealing with natural raw materials, this is where the story gets interesting. Even with this level of complexity, and where there are many variables, AI can find an optimization strategy.  

“All of this doesn’t have to happen overnight,” says Noel. “And it doesn’t have to mean a completely new plant. We can also start by retrofitting older equipment – that too is a first step on the journey to the fully-fledged SmartMill.” 

At its heart, the miller’s expertise 

Although artificial intelligence is fueling this – meaning that a computer system performs tasks that normally require human intelligence – that does not mean that human intelligence is out of the picture. In fact, the approach taken by Bühler keeps human intelligence and expertise at the very heart of everything. 

Rather than developing solutions in isolation, Bühler has worked closely with millers for years to identify real pain points and define what success looks like. “Our approach has always been to work closely with customers. We started with co-creation workshops to explore what is possible and what is important to them. From that, we identified the areas to address, and then we developed solutions together,” says Noel. “It's always a close collaboration between the data scientists, the technologists, and the customer to build something that makes sense from a process point of view, that works technically, and that provides a real benefit.” 

Far from adopting new technology for technology’s sake, Bühler’s approach to implementing AI in milling begins by working together with millers to determine in each case which tasks can be automated, which ones should remain with the human operator, and which can be a combination – where AI augments the human. “We approach this by asking what value does it deliver? We look for the use cases where AI makes sense; where it can augment human judgement rather than replace it,” says Noel. 

The resulting system also allows for more transparency. SmartMill tracks all the metrics the miller wants to monitor and all the changes in the process configuration. “This is not a black box operating on its own. Human operators can check the input data that are used to build it,” says Noel. 

Equally important is the application of safety guardrails. Powerful technology demands responsibility. For Bühler, responsible use of AI means being transparent about where and how AI is applied, protecting personal and business-sensitive data, validating outputs before acting on them, and keeping humans accountable for the decisions that matter.  

In practice, this means Bühler uses its own AI models and trains them in-house with data that is 100% relevant for the task. All the operations Bühler carries out to develop or maintain software are externally audited and ISO certified. “We maintain the highest standard in cyber security, and we are ready for the introduction of the Cyber Resilience Act in the EU. We have developed a secure link between the cloud and plants so that we can deploy security patches,” says Noel. 

Every change is also a cultural change 

Safety is also enhanced by Bühler’s step-by-step approach to the SmartMill. “We don’t do it all at once. It starts with one recipe, one passage, one roller mill, and then it expands. This makes it easier for operators to adapt,” says Schafler. “We are not asking anyone to go straight from driving a manual car to letting an autonomous vehicle do the driving. We know that trust has to be built up.” 

This does entail a change in working habits, but the change can happen with the people at their pace and with the right support. “As we move away from pushing buttons to monitoring screens, it requires a mindset change for operators and management alike,” says Schafler. “To help with this, we provide training and support to use these tools and functions. It is an investment in time, but it is important to take these first steps, rather than skip them and try to jump in at the final stage.” 

Bühler takes training equally seriously in-house. Its apprenticeship programs are embracing AI with a focus on questions of responsibility, transparency, fairness, and data protection. The next generation is already being trained to think about who bears responsibility for the decisions and consequences of AI systems, and how a conscious and secure approach to AI can be developed.  

“AI is a collaboration between humans and machines taken to the next level. If we do this right – choosing the right tasks to tackle, using the right data, and with the right people in charge at the right points in the process, there is much to be gained,” says Schafler. 

The SmartMill is a journey of transformation, not a disruption.

Georg Schafler,
Head of Wheat and Rye Process Technology at Bühler 

Relying on unique expertise 

Since AI went mainstream in 2022, people everywhere have started experimenting with the technology – at home, at work, and at school. “It has become so commonplace that the younger generation already takes AI for granted,” says Noel. “But building an AI feature into a piece of software is much easier than integrating it into a machine and running it in an industrial process in real time. That is where we stand out – we have the capabilities to take a scientific breakthrough like this and learn how to use it to deliver real value in a traditional industry like milling.” 

Other providers might be developing interesting concepts in this space, but Bühler, uniquely, is able to build on its deep process expertise and knowledge of the industry to embed AI in its solutions and create a seamless interaction between the equipment and the software. “Because we provide equipment, process expertise, and services, we occupy a central position in the whole ecosystem,” says Noel. “With Bühler’s solutions, you know that everything is aligned, from the machine to the software – all with the same quality. And we promise that the development will not stop. We continue to add further functionality as the AI frontiers expand.” 

 “We have worked closely with our customers in some cases for generations and know their pain points very well. We know that millers want to increase capacity, efficiency, and quality, and reduce costs. We know that in many places they face a skills shortage, rising energy prices, and increasing fluctuations in the raw materials. And we know that the ability to make adjustments across a number of variables, across machines, lines, and plants in real time can mean winning on margins and staying competitive. We develop solutions to these challenges with our customers and accompany them on every step of this journey,” says Noel. 

“We also know that running machines is not the primary goal of a miller. Millers are in the business of grain processing – in the art of flour production – their expertise is in getting the right raw materials and developing the right recipes to produce the products their customers want. Maintaining equipment and knowing how to adjust it is just an enabler of this. Our aim is to support them with this, so that millers around the world can achieve their goals.”  

BACK TO THE INSPIRATION HUB