IQP – Intelligent Quality Platform

Increasing productivity with data analysis and AI methods

AI applications and machine learning promise manufacturing companies greater productivity and increased efficiency. Examples of this include more precise maintenance scheduling (maintenance prediction), optimized production program planning and more reliable quality controls. Prototype applications show promising results, but are often not sustainably integrated into the production infrastructure and offer no long-term added value for users. Prototypes cannot be scaled and are ultimately uneconomical.

The Intelligent Quality Platform (IQP) bundles these modern applications under one roof. From anomaly detection to visual quality control, the IQP is a software platform for deploying, monitoring and continuously optimizing industrial machine learning (ML) applications. The IQP enables efficient and economical use of different ML applications from different production areas through standardized integration, monitoring and parallel operation. The platform uses standardized structures to minimize the additional effort required to recreate already developed processes for individual AI and data analysis projects.

Benefits at a glance:

  • For management: Overview of different ML applications
  • For production. Machine learning improves efficiency on the shopfloor
  • For software development: Practical insights for successful ML implementation

Applications in the industry

Industrial project

Labormuster für die Produktion von Infrarotoptiken aus Chalkogenidglas im Wafer-Maßstab (100 mm)

The contract manufacturing of passive cooling components based on aluminum nitride plays an important role in the electronics industry. This production involves the precise application of coating patterns using photolithography and the subsequent separation of the individual heat sinks from the wafer.

Various defects can occur during production, including defects in the applied patterns, edge damage to the heat sinks due to problems during the separation process and mechanical damage that can occur during production.

The challenge

Detecting these defects through a full manual visual inspection is error-prone and subjective, which can lead to a high false positive rate.

Goal

The goal of this project is to implement an automated visual inspection that offers several key benefits. By minimizing the false positive rate, more accurate defect detection is achieved, leading to a significant reduction in false alarms.

Economic impact

By optimizing production processes and reliably detecting defects, the following benefits can be achieved:

  • Highest product quality: avoiding the delivery of faulty products.
  • Efficient production: Continuous improvement of production processes based on the analysis of detected faults.

Learn more about the results of this project and numerous other use cases on our Intelligent Quality Platform. Make an appointment now and get to know the IQP.

Research project

Glasmikrolinsen und Presswerkzeuge

In this use case, machine learning is used to predict the average thickness of precision glass produced by precision glass molding (PGM). Existing machine data such as temperature and pressure are used. The aim of the developed algorithm is to automate quality assessment, reduce manual measurements and set up an alarm system for products that do not meet specifications in order to increase overall production efficiency.

The challenge

Integrating different machine data sources, ensuring data quality and consistency and developing a robust center thickness prediction model are the main challenges. In addition, selecting relevant information from the machine data is essential for model training and maintaining model accuracy in a dynamic manufacturing environment with changing conditions.

Objective

The main objective of this project is to predict the mean thickness of lenses during the precision glass molding process by machine learning. By analyzing real-time machine data, accurate and timely predictions will be made, enabling proactive quality control and reducing reliance on manual measurements.

Economic impact

The implementation of predictive algorithms has the following economic effects:

  • Cost reduction: lowering production costs by reducing manual measurements.
  • Quality improvement: Minimizing the production of glass that does not meet specifications.
  • Customer satisfaction: Improving customer satisfaction through consistently high-quality glass production.

Learn more about the results of this project and numerous other use cases on our Intelligent Quality Platform. Make an appointment now and get to know the IQP.

Get an insight into the IQP

The IQP enables manufacturing companies to sustainably increase their productivity through innovative methods of data analysis and artificial intelligence in various use cases. With the help of a variety of tools, all aspects of the use case can be quickly and easily examined.

Dashboard

On the dashboard, the IQP provides a compact description of the use case, including the associated challenges, objectives and economic benefits. The dashboard reads out production data in real time and thus provides a direct insight into the quality and productivity of production.

Quality Insights

Interactive diagrams allow you to easily check the quality of production and the machines used. Functions such as "Individual process analysis" and "Machine parameter optimization" are available to you. Additional functionalities provide a detailed insight into the production processes.

Pipeline Management

The applications are categorized according to their functionality in the use case in a clear process structure. This ensures that the required data for the respective ML process can be found quickly and efficiently. Each app also offers detailed insights into the respective process.

IT Resources

All IT resources in use at a glance: Clearly sorted by CPU, memory and hard disk, IQP users receive all the necessary information about the utilization status of their IT resources.