PROSAM - Anticipate Faults before they occur

Knowledge and data-based approaches improve fault prognosis models to enable the development of anticipative maintenance strategies, thereby optimizing operational availability and resource efficiency of manufacturing processes.

Short Description

Anticipative maintenance strategies bear various benefits:

  • An increased plant availability, as the early identification of faults can reduce down-times;
  • a reduction of material and energy costs, since maintenance activities are no longer carried out on pre-defined schedules but only when necessary;
  • and an easier planning of maintenance procedures, due to continuous condition monitoring.

Despite this huge optimization potential and the availability of more and more condition-relevant process and sensor data, anticipative maintenance strategies are only rarely found in industrial applications. Reasons for this are the growing complexity of production plants and the increased diversity of used components. These pose one of the major challenges to the development of a strategic maintenance management.

The project PROSAM, however, rises to this challenge and aims to solve it via the development of novel conceptual as well as methodological foundations, where the combination of expert knowledge and data based fault prediction models is considered as the key factor to success and therefore strongly pursued by consortium leader SCCH.

Within this process, it is crucial to incorporate all critical aspects from

  • data integration and processing,
  • over feature extraction,
  • model building,
  • knowledge representation and
  • problem oriented system analysis up to
  • the optimal integration of developed techniques within the maintenance management.

As an expert on condition monitoring, project partner Messfeld GmbH provides the know-how required to generalize component-oriented monitoring approaches to plant-wide strategies.

H&H Systems, the third consortium member and producer of a maintenance management tool, plays a key role in developing a methodology to integrate prognosis models, which are naturally subject to a certain amount of uncertainty, inside a comprehensive maintenance plan. This step is particularly challenging with respect to the actual practical applicability, as various constraints regarding economical as well as operational planning aspects need to be taken into account.