The pharmaceutical and consumer goods industries in particular work with capital-intensive production facilities and are dependent on maximizing productivity at all times. Otherwise, there is a risk of cost pressure and financing gaps, as well as the fact that many production systems comprise a large number of stations and work so quickly that the causes of errors cannot be detected with the naked eye.

In 'smart system optimization', which Felix Müller, Group Manager Autonomous Production Optimization at Fraunhofer IPA, has developed together with his team, a powerful connector accesses the data in the machine control system at high frequency via the respective manufacturer protocol. This creates a continuous database that several self-learning algorithms evaluate synchronously. These recognize in detail where faults are present in the production system, how they are connected and what priorities they have for rectification. In this way, defects that lead to the failure of the entire system can be rectified more quickly or even predicted.

Shannon - an intelligent worker assistance system

However, it is not always clear what to do if an error threatens to occur. In addition, there are follow-up messages from the system that make the situation even more confusing for the operator. For this reason, Müller and his team have developed 'Shannon', an intelligent operator assistance system for complex production systems that is based on smart system optimization. Previously, it was often up to the machine operators to decide what to do to rectify a fault. Now, however, the affected machine sends them detailed step-by-step instructions on their smartphone or tablet. The database and the links between faults and solutions are constantly expanding during system operation.

This gives plant operators the opportunity to create their own instructions, for example for rectifying faults. In addition to text, these instructions can also include photos and videos. The system operator can also provide feedback on the information provided, which is used to improve it. System operators are also actively encouraged to contribute their knowledge, for example when describing detected but unknown events. In this way, a clearly understandable and consistently linked knowledge database is built up over time, consisting of faults, events and solutions. Shannon is currently being used as a tablet and smartphone app in several factories, where it has significantly reduced the time it takes to rectify faults.

Benchmarking increases efficiency by up to 18 percent

Automated machine benchmarking is also feasible with smart system optimization: In many production halls, there are dozens of identical or similar machines that always carry out the same processing cycle. One example of this is injection molding, die casting, blow molding and thermoforming machines. Although they all have the same design, some work slower than others. This is usually due to wear on certain components, varying sensor behavior or different tool settings and material fluctuations.

In machine benchmarking, the overall process in a machine is first defined and broken down into individual steps. The high-frequency connector on the machine controller then generates a database that is evaluated by a machine learning algorithm package. This happens simultaneously with all connected machines and is merged virtually into an ideal process sequence. From this, the tool immediately recognizes when a machine is running slower than intended and links this to a technical cause. Users can thus not only rectify faults before they occur, but also achieve an optimized cycle time for the connected machines by merging the best individual steps. Depending on the machine, this has led to cycle time reductions of between two and 18 percent in previous prototype applications. The application has now been transferred to a continuously learning software called Darwin.