Granularity

Granularity refers to the level of detail or resolution at which data, processes, activities, or controls are defined, recorded, and analyzed. In industrial and regulated manufacturing environments, it describes how finely information or operations are broken down into smaller elements.

Granularity in manufacturing and operations

In manufacturing systems, granularity commonly applies to:

• Production data: Level of detail for tracking (e.g., by plant, line, work center, work order, operation step, or individual cycle).
• Traceability records: How specifically materials and components are linked to finished goods (e.g., lot-level vs. serial-number-level, or down to a specific operator and shift).
• Process definition: How detailed routings and work instructions are (e.g., a single generic operation vs. multiple discrete steps with parameters and checks).
• Quality and NC data: How nonconformances, defects, and measurements are captured (e.g., by batch vs. by individual characteristic or feature).
• Event logging and audit trails: Degree of detail for system and user actions (e.g., only major status changes vs. every field change and timestamp).

Higher granularity means more detailed, smaller units (such as individual parts or operations). Lower granularity means more aggregated units (such as batches, orders, or time buckets). The appropriate level of granularity is usually chosen to balance traceability, analysis needs, and system complexity.

Operational examples

• An MES configured to record scrap at the operation and serial number level has finer granularity than one that records scrap at the daily work-order level.
• A quality system that logs measurements for each ballooned characteristic individually uses higher data granularity than one that stores only a pass/fail result per part.
• Production performance dashboards can show OEE at coarse granularity (per plant per day) or fine granularity (per machine per hour or shift).

Common confusion

• Granularity vs. frequency: Granularity is the level of detail; frequency is how often data is collected. Data can be sampled frequently but still have low granularity if it remains highly aggregated.
• Granularity vs. precision: Granularity describes how finely information is segmented (e.g., per serial number). Precision describes the exactness of a value (e.g., number of decimal places in a measurement).

Relation to digital systems

Granularity is a key design decision for MES, ERP, PLM, QMS, and data integration. It affects:

• How easily root cause analysis, genealogy, and audit evidence can be produced.
• How complex master data, work instructions, and routings become.
• Data volume and performance in reporting and analytics tools.

Defining appropriate granularity typically involves cross-functional input from operations, quality, IT/OT, and compliance stakeholders.