Birgit Vogel-Heuser


Technical University of Munich

Modular and adaptive field level automation architectures to support predictive maintenance

Field-level automation architectures often provide limited access to information beneficial for predictive maintenance due to security, real-time, and reliability concerns. On the other hand, it is neither possible nor reasonable to make all field level information available due to network limitations and cost. Nevertheless, the Digital Twins of sensors, actuators, units and other levels of ISA 88 need to be updated and evolve with information gained during operation. The talk will present the existing field-level automation architectures including their weaknesses and strengths regarding their evolution and the evolution of their Digital Twins. Particularly in special purpose machinery and plant manufacturing, there are often too many field-level alarms to be beneficial for predictive maintenance. Consequently, the pre-processing and reduction of such alarms will also be addressed.

Dimitrios Kyritsis

Professor Emeritus

École Polytechnique Fédérale de Lausanne
University of Oslo
Ontology Based Asset Information Modeling for Predictive Maintenance

In this keynote talk, I will present our experiences gathered in various European projects such as BOOST4.0, Z-BRE4K, RE4DY and SM4RTENANCE where we are developing ontology-based methodologies and tools for predictive maintenance. In this context, we have developed a pilot for the manufacturing of spindles of machine tools. This pilot includes an Extract-Transform-Load pipeline to collect data from heterogeneous sources, transform it by adding semantics, and store it as linked data, using the latest software development technologies. It also uses an ontology for the semantical model, in order to standardize data and ensure that data is expressed using one common model all across the factory. The data can then be queried, visualized and also analyzed, because it is machine-readable thanks to the semantics, and an algorithm of machine learning runs on it to deliver information for predictive maintenance. Through this pilot, we demonstrate the possibilities that the combination of ontologies and modern software architectures can offer in terms of asset digitalization, data standardization and data analysis, indicating a path towards Factories 4.0.

Adolfo Crespo Marquez


Universidad De Sevilla

Lessons from the GFMAM 25DX Project: Unlocking digital transformation in maintenance and asset management through a global initiative

The Global Forum on Maintenance and Asset Management (GFMAM) is a collaborative hub for knowledge and standards exchange in the maintenance and asset management domains. As the digital transformation (DX) landscape rapidly evolves in this field, presenting both challenges and opportunities, this paper delivers a comprehensive overview of key findings, implications, concerns, and guidelines to aid in the facilitation of this transformative process, according to Project 25 DX results. The analysis emphasizes the profound impact of DX on asset management and maintenance, highlighting the critical roles of data, analytics, compliance, optimization, human resources, and holistic lifecycle management. Encouraging organizations to embrace change, foster innovation, and adapt strategies and processes, the paper also reflects community concerns through an extensive survey of industry professionals across diverse sectors and roles. The survey identifies prevalent challenges, barriers, and readiness for digital transformation, offering tailored guidance and recommendations for professionals and organizations at various levels of asset management maturity. This work provides an initial glimpse into the upcoming GFMAM document on DX in maintenance and asset management, slated for publication by early 2024.