Data mining and data defect prevention

Challenge

In the course of rework activities in production, a large amount of data must be entered that provides information about the exact rework activities. Some of this data is entered manually, which leads to shortcomings in terms of completeness and content quality.

Further data evaluations and visualization to identify necessary improvement measures in production thus forfeit quality and informative value.

Solution

Using selected methods from the field of machine learning, analyses are performed on existing data to determine correlations and relationships between the data.

These findings are used to derive concrete recommendations for action that help to improve data quality – generally referred to as Data Defect Prevention.

Benefit

• A proposal function of the AI can achieve faster data acquisition during defect entry.
• Data maintenance can be improved using dynamic plausibility checks.
• Reports and evaluations can be improved by an automated, retroactive addition of data and incorrect or implausible data can be marked accordingly.