Reduction of the error rate and rejects through real-time analyses and forecasts during operation

Quality control is essential in every value-added process and especially in production. The aim is to avoid errors by using predictions as a basis for decisions on measures – and thus reduce rework, costs and additional work.

The problem today: error sources and indicators are often only identified retrospectively when the costs have already been incurred.

Reduction of the error rate and rejects through real-time analyses and forecasts during operation

Quality control is essential in every value-added process and especially in production. The aim is to avoid errors by using predictions as a basis for decisions on measures – and thus reduce rework, costs and additional work.

The problem today: error sources and indicators are often only identified retrospectively when the costs have already been incurred.

ASCon combines the Digital Twin technology with its quality management know-how

The Predictive Quality Assistant enables you to monitor and continuously analyze the essential process parameters in the context of your entire value-added process – and thus identify potential problems before they arise. The Digital Twin enables companies to perform data-driven, AI-based monitoring of product and process-related quality and supports the user in diagnosis and decision-making.

Collect
Data

Signals, parameters and information are recorded in a holistic data context and form a “smart” database.

Analyze
Data

Analysis of live and historical data from machines, sensors, ERP, MES and quality systems to detect anomalies.

Predict
Quality

Forecasting of quality by identifying and monitoring process, product and resource parameters that previously led to errors.

Alert
Teams

Early notification of plant managers, operators and quality assurance of potential quality defects with appropriate lead time.

Process & Product Quality measurements

  • Monitoring of production processes at plant, line and factory level
  • Determination and monitoring of individual process statuses
  • Detection of anomalies, which can be relevant to NIO, for example
  • Display and search for anomalies and error messages on the basis of historicized and contextualized data in the digital file (live and retroactive)
  • Detection of anomalies during ongoing process operation

Predictive Quality Analysis

Predictive quality analysis is based on artificial neural networks. The spectrum of analysis methods includes both statistical methods and methods of machine learning including deep learning. Here we integrate AI services of our technology partner Synergeticon, whose analyses support quality assurance and determination of production errors by evaluating 3D and sensor data.

 

The following analyses and functions are among them:

  • Algorithms for regression
  • Cluster analyses
  • Trend and time series analysis
  • Machine Learning

Predective Quality Analysis

Predictive quality analysis is based on artificial neural networks. The spectrum of analysis methods includes both statistical methods and methods of machine learning including deep learning. Here we integrate AI services of our technology partner Synergeticon, whose analyses support quality assurance and determination of production errors by evaluating 3D and sensor data.

 

The following analyses and functions are among them:

  • Algorithms for regression
  • Cluster analyses
  • Trend and time series analysis
  • Machine Learning

Linkage of product, process and resources

Using the data-based approach of the Digital Twin, all relevant parameters, sensors and environmental data are first recorded and standardized in a semantic model.

In addition to online data acquisition from existing IT systems, offline data sources are also relevant, such as quality messages from downstream processes. The data is integrated in the Digital Twin and is then provided to the user for data storage, analysis functions and for decision preparation or visualization.

The value of predictive quality is not in the data itself, but in the knowledge generated from the context of product and process quality, which flows directly into decision support.

Do you have questions or are you interested in a product demonstration?

Your contact person

Tilo Geisel
Tech Sales Manager

Contact us