Yes, a digital thread can be built when different plants use different MES systems. The practical answer is not to pretend the MES landscape is uniform, but to define what data must be connected, how it is identified, where the authoritative records live, and how mappings are controlled over time. In regulated brownfield environments, this is usually an integration and governance problem, not simply a software replacement problem.
A common MES is not a prerequisite for a digital thread. What matters more is whether each plant can expose reliable, traceable data for the product, process, material, equipment, quality events, revisions, and execution history that the thread is expected to connect.
What has to be common
Different MES platforms can coexist if the enterprise establishes common rules for the data that crosses plant boundaries. This often includes:
- Common part, serial, lot, batch, and work order identifiers, or controlled cross-reference rules.
- A canonical data model for the information shared across MES, ERP, PLM, QMS, maintenance, and analytics systems.
- Clear ownership of master data and revision-controlled definitions.
- Consistent handling of timestamps, units of measure, operation names, dispositions, and status codes.
- Audit trails showing when data was created, changed, mapped, approved, or corrected.
If those foundations are weak, the digital thread may still be possible, but it will be partial, fragile, and dependent on manual reconciliation. That may be acceptable for a limited use case, but it should not be confused with an enterprise-grade thread.
Where different MES systems create risk
The main risk is semantic mismatch. Two plants may both record a completed operation, a nonconformance, or an inspection result, but the meaning, timing, approval state, and data granularity may differ. If those differences are hidden behind a dashboard, the thread can produce misleading conclusions.
For example, one MES may record an operation as complete when the operator finishes work. Another may record completion only after inspection or quality buyoff. Both records may be valid locally, but they are not equivalent without mapping and context.
Other common failure modes include incomplete genealogy, inconsistent serial or lot tracking, undocumented local workarounds, weak interface monitoring, and poor change control when routings, inspection plans, or item masters are updated.
Why full MES replacement is usually not the first answer
Replacing every MES with one standard platform may sound cleaner, but it is often unrealistic in aerospace-grade and similarly regulated operations. The qualification burden, validation cost, downtime risk, interface complexity, operator retraining, and long equipment lifecycles can make a full replacement program slow, expensive, and operationally risky.
In many plants, MES is also tied to local machines, test stands, inspection equipment, label printers, maintenance systems, legacy ERP interfaces, and customer-specific reporting. Removing it can disturb more than production execution. A coexistence strategy is often more credible, provided the integration layer and governance model are treated as controlled systems, not informal IT plumbing.
What a workable approach looks like
A practical digital thread across mixed MES environments usually starts with a bounded scope. The first thread may focus on a product family, a critical part, a customer program, a compliance record, or a failure investigation workflow rather than every data element in every plant.
The work typically includes mapping source systems, defining authoritative records, reconciling master data, documenting business rules, validating interfaces, and placing changes under formal control. Where the thread supports regulated records or customer deliverables, testing and evidence retention matter. The system may improve traceability, but it does not by itself guarantee audit acceptance, customer approval, or regulatory compliance.
Bottom line
Different MES systems do not prevent a digital thread. Poorly governed data does. The question is whether the organization can control identifiers, definitions, interfaces, and change processes across plants. If it can, a federated digital thread is realistic. If it cannot, the result will likely be a collection of integrations that require manual interpretation when the data is challenged.