Repository logo
 
No Thumbnail Available
Publication

Integration of machine learning models in a microservices architecture

Use this identifier to reference this record.
Name:Description:Size:Format: 
Ahmed Ibrahim.pdf3.88 MBAdobe PDF Download

Abstract(s)

Achieving Zero Defect Manufacturing in the evolving landscape of Industry 4.0 requires advanced, scalable architectures that support proactive quality management and real-time defect detection. This thesis introduces a ZDM-focused microservices architecture designed to enhance modularity, resilience, and scalability within industrial manufacturing environments. By integrating Cyber-Physical Systems and Digital Twins, the proposed architecture facilitates continuous monitoring, dynamic data flow, and predictive analyt- ics, aligning with RAMI 4.0 standards to ensure seamless interoperability across systems. Emphasizing communication-driven design, the architecture leverages distributed microservices and specialized communication brokers to create a flexible, event-driven system. This enables efficient handling of high data volumes, real-time quality insights, and early anomaly detection. Through a structured evaluation of core architectural components, including orchestration, choreography, and communication brokers, this work establishes a foundation for ZDM implementations adaptable to various industrial settings. The architecture’s deployment in a real-world manufacturing case demonstrates its practical benefits, illustrating how modular, scalable systems can drive operational improvements and defect reduction. Contributing to the broader Industry 4.0 framework, this work provides a blueprint for future ZDM solutions that prioritize sustainability, adaptability, and enhanced product quality in complex manufacturing ecosystems. Keywords: Microservices architecture, Message brokers, Industry 4.0

Description

Keywords

Microservices architecture Message broker Industry 4.0

Citation

Research Projects

Organizational Units

Journal Issue