Glossary

Integration boundary

The defined limit between systems, processes, or data domains where information is exchanged and responsibility changes.

Data Mapping and System Interoperability ERP / MES / PLM Interoperability

An integration boundary is the defined limit between two systems, applications, process areas, or data domains where information, events, or control signals are exchanged. It marks where one system’s responsibility ends and another begins.

In manufacturing and regulated operations, the term commonly refers to the interface between platforms such as MES, ERP, PLM, QMS, LIMS, SCADA, or shop-floor equipment. The boundary may include data structures, message formats, timing rules, ownership of records, validation checks, security controls, and error-handling rules.

An integration boundary is not the same as the integration itself. The integration is the mechanism or workflow used to connect systems. The boundary is the line that defines what crosses between them, under what conditions, and which system is authoritative for each type of data.

What it typically includes

The systems or process areas on each side of the interface
The specific data, documents, events, or transactions that cross the boundary
Source-of-record ownership, such as which system owns part masters, work orders, production status, or quality results
Trigger points, frequency, and timing, such as real-time, near-real-time, or batch exchange
Validation, exception handling, and reconciliation rules
Access, security, and audit-related constraints where applicable

How it appears in operations

Operationally, an integration boundary often appears when a business process moves from planning to execution, from execution to quality review, or from plant systems to enterprise reporting. For example, ERP may release a production order to MES across an integration boundary, while MES returns completion quantities, labor, material consumption, and traceability data back to ERP.

Clear integration boundaries help teams describe which records are created in one system, which are only referenced, and which are updated across systems. This is especially important where data integrity, version control, traceability, and controlled changes matter.

Common confusion

Integration boundary is commonly confused with system boundary. A system boundary describes what is inside or outside a single system’s scope. An integration boundary focuses on the handoff or exchange point between systems.

It can also be confused with an API or interface. An API or interface is a technical means of connection. The integration boundary is broader and includes business responsibility, data ownership, and process rules, not just the transport method.

It is also different from a network boundary in cybersecurity. A network boundary concerns segmentation and communications control between networks or zones. An integration boundary concerns business and application-level exchange, though the two may overlap in OT and IT environments.

In aerospace plants, ERP plans and accounts for the business (orders, MRP, costing, finance), while MES controls and records execution on the shop floor (work instructions, data collection, traceability). They must coexist and integrate; trying to make one system do both usually breaks under aerospace compliance, validation, and traceability needs.

What integration standards pair well with ISO 22400 in aerospace manufacturing systems?

ISO 22400 defines KPI semantics, not a complete integration architecture. In aerospace manufacturing, it pairs best with standards for equipment connectivity, enterprise-control integration, quality data, and product/manufacturing information exchange. The right mix depends on brownfield realities, data governance, validation scope, and how much semantic mapping your existing MES, ERP, PLM, QMS, and historian stacks can support.

Which data sources should feed an aerospace compliance dashboard?

An aerospace compliance dashboard should pull from the systems that generate controlled records, not from a standalone reporting layer alone. In practice that usually means QMS, MES, ERP, PLM, document control, training, supplier quality, calibration, and audit systems, with careful attention to record ownership, data quality, timestamp alignment, and validation.

How does a unified KPI framework enable predictive analytics and AI?

A unified KPI framework helps predictive analytics and AI only if it standardizes metric definitions, timestamps, context, and data lineage across systems. It improves model training, comparability, and trust, but it does not fix poor source data, weak integrations, or immature processes. In brownfield plants, value usually comes from governed coexistence, not full platform replacement.

Related Glossary

brownfield environments

Existing industrial sites or systems that are already built and operating, where new equipment, software, or controls are added or modernized.