Predictive maintenance in industry workshop
Location and timing
The POM consortium is organizing an workshop around the theme: Predictive Maintenance in the Industry
The workshop will take place on 30th of March, between 14-17.30uur, in Gent, Technologiepark 915, BE-9052 Zwijnaarde
14.00 - Welcome 14.10 - Predictive maintenance in practice, Pieter Van Camp 15.00 - On-line condition monitoring of bunching equipment, Carlo Cloet, Bekaert 15.30 – Who says the mechatronic industry is dead? Verhaert 17.00 - How to determine your optimal maintenance policy, Adriaan Van Horenbeek, CIB/KUL 17.30 - Status of POM project, Steve Vandenplas 17.45 - Discussions and Drinks
Dr. Ir. Steve Vandenplas - Monitoring and Diagnostics program manager FMTC The Monitoring and Diagnostics project cluster at FMTC focuses on two primordial competences: monitoring solutions and fault diagnostics.
- Monitoring solutions: In some cases, commercial available sensing and condition monitoring solutions are currently non-existing or too expensive to be integrated into machines. For dedicated problems, FMTC is looking towards alternatives sensing technologies or methods evolving advanced signal processing or virtual sensing.
- Fault diagnostics adapted to the needs of the machine builders. A fault is an unpermitted deviation of at least one characteristic property of the machine’s normal standard condition. A fault may initiate a failure or malfunction. Within this competence we focus on fault detection, fault identification, and fault diagnostic modeling (modeling the relationship between symptoms and the fault). Hereby signal-based, data mining based and physical model based methods are developed for a wide range of industrial applications.
Furthermore, in 2009 FMTC started a long-term research activity on smart self-diagnostics. Self-diagnostics is the machine’s process to find, identify and analyze faults. It is our vision that on-board diagnostic systems of the future are smart. A smart system is capable of describing and analyzing a situation, and taking decisions based on the available data in a predictive or adaptive manner, thereby performing smart actions. Smart self-diagnostics will lead to earlier detection and more accurate identification of faults
Verhart - helps companies and governments to innovate. We design products and systems for organisations looking for new ways to provide value for their customers. Verhaert is a leading integrated product innovation company, creating technology platforms, developing new products and business in parallel, hence facilitating new-growth strategies for our clients.
Pieter Van Camp - Key Account & Technology Manager at Coservices International Coservices is a leading predictive Maintenance Technologies Company. Its purpose is to provide and implement Reliability Based Maintenance products, services and training to enhance industrial profitability and to provide innovative maintenance focused solutions. The company has build a large experience in condition monitoring and diagnostics using vibration analysis, ultrasound analysis, oil analysis and thermography. It offers consultancy services for setting up predictive maintenance programs taking into account the cost benefits analysis for each specific case.
Carlo Cloet - Design engineer in the automation department of Bekaert engineering Carlo is responsible for various automation related projects. His current focus includes algorithm development for fine wire winding, software develoment processes, automated handling systems for factory automation and on-line condition monitoring. Carlo coordinates all joint research efforts with FMTC.
Adriaan Van Horenbeek - PhD student Center for Industrial Management The objective of this doctoral research is to develop a maintenance decision support system that determines business specific optimization criteria, in a first stage. In a second stage a maintenance optimization model (especially focused on condition-based maintenance) will be developed, taking into account the previously determined optimization criteria as well as the supplier/customer relation. The final step will be the exploration of different maintenance service business models. The result of this doctoral research will be a maintenance optimization method both useable in practice and adaptable to the specific business needs of supplier and customer.
POM project description
POM has the ambition to look at the whole maintenance process and opportunities to improve and market maintenance. The tools and methods developed in POM will support the machine builders to follow the present tendency in the industry of moving from pure manufacturing to a performance provider, helping them to maintain reasonable profit margins and making them much more resilient against external competition and delocalization.
At present there is a very delicate ecosystem of suppliers, distributors, maintenance companies and clients around machine building companies. If anything is to change with respect to maintenance, all parties should see a clear advantage. POM has been designed with this goal in mind. Its main objective is to study the whole problem of making a business out of maintenance for the sector of industrial machines and make the advantages of implementing a predictive maintenance program visible.
Concretely the project will:
- DEVELOP DATA AND PHYSICAL MODELS THAT WILL PREDICT THE EVOLUTION OF INDUSTRIAL MACHINE COMPONENTS DEGRADATION, PRODUCT QUALITY AND ENERGY CONSUMPTION IN TIME
- DEAL WITH MACHINE VARIABILITY, NON-STATIONARITY AND RANDOM EVENTS - DATA ACQUIRED FROM MACHINES THROUGH CONDITION MONITORING WILL BE ANALYZED USING MULTIVARIATE STATISTICS AND RECURSIVE PRINCIPAL COMPONENTS BASED ON A TIME VARYING FORGETTING FACTOR
- MAKE THE MOST OUT OF THE MEASURED DATA - DETECT EVENT PATTERNS IN MEASURED DATA BY INCORPORATING EXPERT KNOWLEDGE AND STRUCTURAL RELATIONS BETWEEN MACHINE COMPONENTS
- COMBINE AND AUGMENT PHYSICAL AND DATA DRIVEN MODELS - MODELS WILL COMBINE STATE OF THE ART DATA DRIVEN TECHNIQUES LIKE SUPPORT VECTOR MACHINES AND BAYESIAN NETWORKS