Intelligent Fault Management

The space industry has many similarities with the German high-tech industry in terms of its structures and processes. What these industries have in common is that the series size is small, but the complexity of production is disproportionately high. For these reasons, disruptions play a significant role in the start-up of production. Based on the so-called maximum network plan, a new method for prioritising and solving malfunction messages is to be developed. In addition to the technical target process, the maximum network plan also describes alternative procedures and contains edge attributes that make it possible to weigh these against each other in terms of their quality and any resulting additional effort. This allows decision-makers to always concentrate first on solving the problems that have the highest impact on production progress. Furthermore, modern methods of data evaluation such as data mining are to be used to draw conclusions about potential causes of faults. If solutions have already been found for similar faults in the past and successfully implemented or rejected again, this information should in turn be evaluated using machine learning methods, so that a growing portfolio of ready-made solution proposals including a qualitative evaluation is created for each newly created fault. The concept described is to be implemented and tested as an example in a factory that is in the process of production start-up, the Bremen plant of the ArianeGroup.