Unburdening experts – a lever against labour shortage
15 Oct 2025
The Fraunhofer Institute is developing user-friendly interfaces for untrained personnel – and leveraging the knowledge of experienced employees at the same time. Here, we speak to Raphael Hägle at the Fraunhofer Institute for Manufacturing Engineering and Automation in Stuttgart.
Raphael Hägle, Deputy Head of Research Team Intelligent Manufacturing Processes and Interaction Systems, Fraunhofer Institute.
Intuitive, language-independent user interfaces offer meat product manufacturers many advantages when it comes to reducing the workload of skilled workers by transferring more tasks to unskilled workers. What is your approach to developing and integrating human-machine interfaces (HMIs)?
Our work focuses on thoroughly understanding what the tasks and needs of an operator at an (automated) machine are, and how we can support their work in a targeted way. The operating concept we develop is crucial here, and we consider both the front and back ends. We rely on task-based operating logic, which can be implemented using new technologies like augmented reality or the machine’s existing operating system, such as an intuitive touch-based panel. In my view, the user’s perspective is always central. When recording user tasks and roles, we use eye-tracking systems to analyse gaze data. The HMI, where operation then takes place, should be designed from the user's perspective rather than the developer’s. Needless to say, a human-centred approach should be adopted, with users directly involved in the development process.
What does that entail precisely?
Machines are often built from the developer’s perspective, with the operating system designed accordingly. In practice, however, operators often use machines differently or think in a different way; we see this time and again in the field. We consider these different perspectives and explore the motives together before tackling the redesign. The task-based operating concept focuses on the user’s task, not the functionality of the machine. This forms the basis for redesigning and adapting the HMI.
Eye-tracking systems can analyse the steps involved in operating a machine. Picture: Fraunhofer Institut
What exactly do you offer companies?
For several years now, we have been offering companies a ‘Usability Quick Check’, which involves examining a customer’s machine on site. During the evaluation, we work on the machine with actual operators. We use an eye-tracking system, which involves putting special glasses on the users to analyse more closely what they look at while they’re working. Once the evaluation is complete, we can move on to implementation. Our diverse team comprises programmers, designers, and usability experts, enabling us to develop a holistic, user-friendly operating concept. Based on the results of the ‘Usability Quick Check’, we can then address further development and optimisation. We also support companies in developing an HMI from scratch. An exciting possibility here is implementing a web-based HMI architecture, where the visualisation is separate from the control system. This gives HMI development more freedom and visualisation options.
To what extent are companies willing to invest in these approaches?
In the German region of Swabia in particular, we have companies that employ several thousand people. Many of these companies have already adopted a user-centred development approach, recognising the importance of intuitive operation. However, smaller companies are also setting themselves the strategic goal of developing usability expertise and designing their machines accordingly. It is important to ensure that machines with new ‘extras’ remain affordable and deliver benefits in terms of efficiency gains.
The user-friendly operation of the machine can boost employee motivation. Picture: Smederevac
How do employees respond to ‘intuitive’ panels?
For employees, the ability to operate the machine easily and intuitively has many advantages. In interviews and observations with actual users, we always discuss the topic of motivation: is it more fun to operate a machine that is easy to use? We often hear that when someone buys a mobile phone, tablet or Thermomix, they can start using it straight away. With a machine like this, however, users have to struggle through various settings and menus. In a similar vein, employees mention downtime as very annoying, i.e. when nothing happens for half an hour but the clock is still running.
Intuitive operation can minimise machine downtime. Picture: guteksk7
In what ways can downtime be reduced by using more user-friendly machines?
In a reference project, we successfully trained operators to perform a process step that had previously always required an expert to be called in, resulting in waiting times of up to 30 minutes. We therefore created a workflow in which we broke down and visualised the reconfiguration process on a task-by-task basis. The untrained assistant is now guided through the machine reconfiguration process by the touch-based human-machine interface (HMI). There are some machines that had to be reconfigured three to four times a day. Considering the waiting times that would otherwise arise, the HMI-supported upgrade represented a significant increase in efficiency that can have a concrete impact on alleviating the shortage of skilled workers: because it directly unburdens the expert, the skilled worker. The company increases its efficiency and the manufacturer gains a competitive advantage.
To what extent can user interfaces be customised, and how can AI assist with this process?
One goal would be to create a customised, automatically generated user interface for each machine and customer. This would mean you enter the customer order into the system, and the user interface would be compiled in the background, in accordance with the machine’s functionalities and user roles. I believe there is a lot of potential here. Many companies are still put off by this because they fear it will be very time-consuming. However, we are working with companies to reduce this initial effort, so developers won’t have to manually reconfigure each machine anymore but instead have a system that is automated to always generate the correct user interface.
In the field of knowledge management, there are projects where AI is being trained by experienced operators and is constantly learning. If a problem reoccurs, the AI can suggest a known solution or generate a new one. Overall, AI has great potential to provide targeted support for machine operation.
“I see a lot of potential in offering machines that work in any language and culture. This would allow people who are illiterate to operate machines.”
In Germany, we talk a lot about the importance of craftsmanship and good training. How does this fit in with the integration of unskilled workers?
It all starts with the job at hand: What are the main tasks to be carried out with the machine and what does this mean for the user role? We must move away from the idea that one person is responsible for everything to do with any one machine, from the setup to quality improvement down to the last millimetre. Instead, experts, who are becoming increasingly scarce in Germany, should be deployed in a targeted manner to utilise their skills and experience for quality improvement and/or troubleshooting. Unskilled workers can then be utilised for simpler tasks, supported by HMI systems or augmented reality.
We always start by looking at the roles involved in operating the machine: in one project, there is a process engineer who deals with the data and the operator who does not have much to do with data. The process engineer needs to know the parameters according to which everything runs. The operator keeps the machine itself running.
Thanks to HMIs that are language and culture-independent, even people who cannot read or write can operate machines. Picture: B4LLS
Do you agree that in markets with a lower level of industrialisation, machines that are easier and more intuitive to operate can have an even greater impact? After all, 80 per cent of machines manufactured in Germany are exported.
I see a lot of potential in offering machines that work in any language and culture. This would allow people who are illiterate to operate machines. In our project, we are focusing on using as little text as possible. One of our goals is to make machines that can be culture-independent.
Overall, the extent of automation depends very much on specific industries and markets. China is making significant leaps in mechanical engineering with a high level of digitalisation, which is starting to put pressure on prices here. However, there are many markets where every euro counts and full automation is not possible.
Even here, there are many machines that still require simple manual processes, such as changing control levers or adding rails.
Could a less automated approach also be interesting?
Yes, designing machines with little automation can be exciting, also for markets where simple and robust solutions are crucial. Even for a less automated machine, I can build an operating system that guides the user through the task with text, images, and animations. And if something can only be changed with a spanner, for example, the machine will display this, too.
Michael Hopp
Author at Foodtech Now! editorial office, who wants to show through his stories that tradition and innovation belong together.
