Abstract:Since high data volume and complex data formats delivered in modern high-end production environments go beyond the scope of classical process control systems, more advanced tools involving machine learning are required to reliably recognize failure patterns. However, currently, such systems lack a general setup and are only available as application-specific solutions. We propose a process control framework entitled Health Factor for Process Control (HFPC) to bridge the gap between conventional statistical tools and novel machine learning (ML) algorithms. HFPC comprises two main concepts: (a) pattern type to account for qualitative characteristics (error patterns) and (b) intensity to quantify the level of a deviation. While the system retains large model generality, allowing a broad scope of potential application areas, we demonstrate its favorable mathematical properties in a theoretical analysis. In a case study from the semiconductor industry, we underline that (a) our framework is of practical relevance and goes beyond conventional process control, and (b) achieves high-quality experimental results. We conclude that our work contributes to the integration of ML in real-world process control and paves the way to automated decision support in manufacturing.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: in modern high - end production environments, due to the large amount of data and the complex data formats, traditional process control systems (such as statistical methods like control charts) are unable to effectively identify fault patterns. Therefore, more advanced tools, especially those involving machine learning methods, are required to reliably identify fault patterns and improve the efficiency and accuracy of process control. Specifically, the paper points out that current systems lack a general - purpose framework, and most solutions are limited to specific applications. To solve this problem, the author proposes a framework named "Health Factor for Process Control" (HFPC), aiming to bridge the gap between traditional statistical tools and new - type machine learning algorithms. The HFPC framework contains two main concepts: 1. **Pattern Type**: used to describe qualitative characteristics (such as error patterns). 2. **Intensity**: used to quantify the degree of deviation. Through these two concepts, HFPC can support various potential application scenarios while maintaining the wide applicability of the model. In addition, the author proves the favorable mathematical properties of HFPC through theoretical analysis and verifies its practical relevance and high - quality experimental results in a case study from the semiconductor industry. ### Main contributions of the paper 1. **Proposing the HFPC framework**: integrating pattern recognition and machine learning methods into the process control system, providing a general - purpose process monitoring framework. 2. **Theoretical analysis**: analyzing the mathematical and statistical characteristics of HFPC, deriving expressions for the expected false positive and false negative rates based on HFPC decisions. 3. **Empirical research**: through a case study in the semiconductor industry, demonstrating the performance and advantages of HFPC in practical applications. ### Conclusion The authors believe that their work is helpful for integrating machine learning into real - world process control and paves the way for automated decision - support in manufacturing. At the same time, the reliability of expert knowledge is crucial for reducing the error rate and overall risk of machine learning methods.

Towards a General Framework to Embed Advanced Machine Learning in Process Control Systems

Machine Learning for Process Control of (Bio)Chemical Processes

DeepFMEA -- A Scalable Framework Harmonizing Process Expertise and Data-Driven PHM

Deep Learning of Latent Variable Models for Industrial Process Monitoring

Predictor-Based Data-Driven Model-Free Adaptive Predictive Control of Power Converters Using Machine Learning

Advance Industrial Monitoring of Physio-Chemical Processes Using Novel Integrated Machine Learning Approach

A Drift Handling Approach for Self-Adaptive ML Software in Scalable Industrial Processes

Capability Indices for Digitized Industries: A Review and Outlook of Machine Learning Applications for Predictive Process Control

A Machine Learning Framework For Data-Driven Defect Detection In Multistage Manufacturing Systems

A Holistic Quality Assurance Approach for Machine Learning Applications in Cyber-Physical Production Systems

Adapting to Changes: A Novel Framework for Continual Machine Learning in Industrial Applications

Applications of Machine Learning in Process Monitoring and Controls of L-PBF Additive Manufacturing: A Review

Advancing Machine Learning in Industry 4.0: Benchmark Framework for Rare-event Prediction in Chemical Processes

Safe contextual Bayesian optimization integrated in industrial control for self-learning machines

Modern Machine Learning Tools for Monitoring and Control of Industrial Processes: A Survey

Framework for Holistic Online Optimization of Milling Machine Conditions to Enhance Machine Efficiency and Sustainability

Multiscale Monitoring of Industrial Chemical Process Using Wavelet-Entropy Aided Machine Learning Approach

Industrial process fault diagnosis based on feature enhanced meta-learning toward domain generalization scenarios

An Adaptive Framework for Optimization and Prediction of Air Traffic Management (Sub-)Systems with Machine Learning

A Framework to Model ML Engineering Processes

A general end-to-end diagnosis framework for manufacturing systems