Methodical assessment of the value of sensory captured data regarding the suitability for influencing manufacturing process goals

Abstract

Pricing and assessing sensory captured data is an essential part of the data economy. In current approaches data are assessed based on individual data quality dimensions describing structured properties of data for downstream analyses. From the domain-specific perspective, an assessment regarding the suitability of data for manufacturing process goals is missing. For this reason, this paper presents a methodology for assessing datasets from the manufacturing perspective using an adapted balanced scorecard. This enables the mapping of manufacturing goals in a single score. The assessment is based on a process model that defines the measurable signals of specific manufacturing processes and represents their interactions. To evaluate the importance of process signals, the number of interactions with quality characteristics is examined. Alternatively, the change in affecting quality characteristics with variation of process signals is determined with a sensitivity analysis. On the application of the methodology, a fictitious dataset of a grinding process is analysed.