Definition
Measures the percentage of production records in the digital thread that have complete, validated data linkages across all systems in the thread.
The Measurement Gap
Organizations investing in digital thread architectures typically focus on connecting systems and establishing data flows. Once the integration is live, the assumption is that data flows continuously. In practice, threads break silently. A PLM design record may not link to its MES production record because a field was left blank during data entry. An inspection result may not link to the shipping record because the systems use different lot identification schemes. A field service event may not trace back to the production batch because the serial number format changed during a system upgrade.
These breaks are invisible to anyone who is not actively auditing the thread. The data exists in each system individually, but the cross-system linkages are incomplete. When an AI analytics platform runs analysis across the thread, it works with whatever data it can find. Missing linkages do not generate error messages. They generate incomplete datasets that produce conclusions based on partial evidence.
Without TIS, organizations have no way to know how much of their digital thread is actually intact. They invest in AI analytics that consume thread data without knowing whether the thread is 95% complete or 60% complete. TIS provides the infrastructure quality measurement that determines how much confidence AI analysis on the thread deserves.
Formula
TIS is calculated by sampling production records and auditing cross-system data linkages:
TIS = Records_with_complete_thread / Records_sampled x 100
Where:
- Records_with_complete_thread = production records where every required system linkage is present and data-validated (count)
- Records_sampled = total production records audited in the measurement period (count)
A score of 100% means perfect thread continuity across all sampled records. Lower scores indicate more breaks. Track by thread segment (PLM to MES, MES to QMS, QMS to ERP, ERP to field service) to identify where breaks occur most frequently.
Key considerations for accurate measurement: "Complete" must be defined per product family because some products do not generate data in every system. Consumable products may not have field service data. Make-to-order products may have different PLM linkage patterns than standard products. Adjust the completeness definition accordingly.
Timestamp misalignments between systems may break the thread logically even when data exists in both systems. Include timestamp validation in the integrity check, not just the presence of cross-system keys. For high-volume production, sample statistically rather than auditing every record. A 95% confidence interval on TIS is more practical than a census. Minimum sample size of 50 records per measurement period for meaningful results.
Proposed formula, pending validation pilots.