Platform for Adaptive Integration of Data-Driven Models and Simulations Into Ultra Short Pulse Manufacturing Systems

The field of simulation and machine learning models for ultra-short pulse (USP) ablation manu facturing has seen significant advancements. These models utilize advanced algorithms and mathe matical techniques to predict ablation during material interaction or optimize process parameters, re sulting in improved precision and efficiency. However, the integration of these systems into produc tion is lacking due to the monolithic design of control software. This design, characterized by tightly coupled components and limited modularity, makes it difficult to incorporate new technologies and updates. Additionally, the computational power required for simulations and machine learning algo rithms exceeds the capabilities of traditional shopfloor computers. To address these challenges, a mi croservice-oriented software design is proposed. Microservices allow for the flexible inclusion of data-driven models by deploying and scaling individual components independently. This enables quick adaptation to changing processes, materials, and business requirements. Moreover, micro services offer the flexibility to use multiple models and algorithms, resulting in a more agile system. However, the infrastructure requirements of a microservice-style architecture are higher compared to a monolithic architecture. The authors will discuss the general architecture, software components, and advantages and disadvantages of controlling USP production machines with a microservice-style ar chitecture.