Checking Data Quality
The procedure Data Centric Deliverable Production (40001-EM-SOP-60087) lays out the process and sequence for when should data quality be checked and by whom.
Category 2: Single-tool data quality checks which are critical to be addressed prior to tool/deliverable publishing, to ensure that poor quality data, that cannot easily be rectified by users, is not flushed through the enterprise (e.g. malformed tag name)
Category 3: Single-tool data quality checks which are non-critical to publish as they would be caught or updated during the normal work process (e.g. is the operating pressure greater than the design pressure).
Category 4: Cross-tool, multi-discipline checks which are non-critical to publish, where the consistency of data is checked both before publishing in the tools and after publishing in SPF (e.g. is the temperature in tool A same as in tool B).
Important: As shown in the diagram, the critical data quality report that MUST be run and errors/issues addressed prior to a tool publishing is the Category.2 PRE Publish. Poor quality data in this report MUST be addressed by the discipline tool champions/experts/users BEFORE publishing.
Category.4 PRE-Publish displays inconsistencies between the tool authoring domains.
Category.4 POST Publish shows inconsistencies between the tool publishing domains in SPF.