PostDoc Project
Title:
Machine Learning for predictive maintenance optimization of reconfigurable manufacturing systems
Dates:
2024/07/15 - 2025/06/30
Student:
Supervisor(s):
Description:
A postdoctoral researcher will be recruited at the CRAN laboratory to engage in the Horizon Europe project MODAPTO (Modular manufacturing and distributed control via interoperable digital twin). This project aims to enhance modular and reconfigurable manufacturing systems with production modules that integrate distributed intelligence through interoperable Digital Twins (DTs), adhering to industrial standards. MODAPTO emphasizes a global production perspective, fostering collective intelligence within modular production frameworks to support module and production line design, reconfiguration, and particularly predictive maintenance.
In this context, the postdoc's main task will be to devise AI-driven predictive maintenance algorithms for Industry 4.0, utilizing diverse knowledge sources. These algorithms will undergo testing on pilot site use cases, including those at SEW USOCOME. As a vital member of the CRAN team, the postdoctoral researcher will actively contribute to the MODAPTO project by engaging in development activities, attending and participating in meetings, drafting deliverables, and showcasing progress and outcomes.
In the manufacturing sector, maintenance is crucial for ensuring systems operate within their intended parameters, primarily through predictive maintenance to forestall failures. The goal of maintenance optimization is to manage key performance indicatorscovering both traditional (like productivity and reliability) and emerging aspects (such as sustainability)efficiently and cost-effectively. The advent of Industry 4.0 introduces new production methodologies and maintenance techniques essential for enhancing the agility and resilience of manufacturing systems. The rise of Cyber-Physical Systems (CPS) and Cyber-Physical Production Systems (CPPS) marks a significant digital shift, generating extensive data on system health and operational conditions. This data, analyzed using advanced AI methods, enhances the detection of equipment anomalies, predicts potential failures, and informs proactive maintenance decisions. However, the use of AI in maintenance and prognostics is not fully realized and requires bespoke solutions for specific production systems. Given the dynamic nature of CPPS, which adapt to new production needs, AI-based predictive maintenance approaches need to incorporate reasoning abilities that understand system structures and constraints, allowing for adaptation to changes.
The objective of this postdoctoral project is to develop AI-driven machine learning and reasoning tools for predictive maintenance in Industry 4.0, aiming to optimize maintenance actions in a dynamic, timely, and context-aware manner for reconfigurable manufacturing environments.
In this context, the postdoc's main task will be to devise AI-driven predictive maintenance algorithms for Industry 4.0, utilizing diverse knowledge sources. These algorithms will undergo testing on pilot site use cases, including those at SEW USOCOME. As a vital member of the CRAN team, the postdoctoral researcher will actively contribute to the MODAPTO project by engaging in development activities, attending and participating in meetings, drafting deliverables, and showcasing progress and outcomes.
In the manufacturing sector, maintenance is crucial for ensuring systems operate within their intended parameters, primarily through predictive maintenance to forestall failures. The goal of maintenance optimization is to manage key performance indicatorscovering both traditional (like productivity and reliability) and emerging aspects (such as sustainability)efficiently and cost-effectively. The advent of Industry 4.0 introduces new production methodologies and maintenance techniques essential for enhancing the agility and resilience of manufacturing systems. The rise of Cyber-Physical Systems (CPS) and Cyber-Physical Production Systems (CPPS) marks a significant digital shift, generating extensive data on system health and operational conditions. This data, analyzed using advanced AI methods, enhances the detection of equipment anomalies, predicts potential failures, and informs proactive maintenance decisions. However, the use of AI in maintenance and prognostics is not fully realized and requires bespoke solutions for specific production systems. Given the dynamic nature of CPPS, which adapt to new production needs, AI-based predictive maintenance approaches need to incorporate reasoning abilities that understand system structures and constraints, allowing for adaptation to changes.
The objective of this postdoctoral project is to develop AI-driven machine learning and reasoning tools for predictive maintenance in Industry 4.0, aiming to optimize maintenance actions in a dynamic, timely, and context-aware manner for reconfigurable manufacturing environments.
Department(s):
| Modeling and Control of Industrial Systems |