Start with a small, stable set of MES scrap and rework KPIs
In aerospace environments, the most useful MES scrap and rework KPIs are those that can be consistently calculated from production data, reconciled with ERP/QMS, and traced back to specific orders, operations, and resources. A small, well-defined set is usually more effective than dozens of weakly-governed metrics. Because of qualification and validation burdens, changing KPI definitions later is expensive and can break historical comparisons, so plants benefit from agreeing up front on calculation rules and data owners. The MES normally surfaces operation-level details (who, when, where, how), while ERP and QMS carry financial and formal disposition data, so KPIs must be designed to bridge these domains. In brownfield plants, practical KPIs are often constrained by what legacy routing structures, defect coding schemes, and data collection methods reliably support.
Core yield and scrap KPIs at operation and order level
First, most plants benefit from a basic set of yield and scrap KPIs defined at both operation and order level. Typical metrics include First Pass Yield (FPY) per operation, overall yield per routing, and scrap rate per work order or serial number, always using clearly defined numerators and denominators. In aerospace, FPY is especially useful if the MES can distinguish between minor rework within the same operation and true rework loops or returns to prior operations. Scrap quantity and scrap rate should be available by part number, revision, work center, shift, and operator where possible, but only if data quality and union or privacy constraints allow that granularity. These KPIs are only meaningful if the MES consistently captures good/defect quantities at the point of use and if routing changes are under tight change control so that historical comparisons remain valid.
Defect and rework profile KPIs linked to traceability
Beyond simple scrap rate, aerospace plants usually need KPIs that describe the defect and rework profile in a way that supports root cause analysis. Useful MES-level metrics include defect rate by defect code, by operation, by work center, and by key material or tooling identifier. Rework rate (percentage of pieces requiring any rework) and rework loops per unit (how many times a unit leaves the standard routing to be reworked) help quantify process instability. These KPIs depend on disciplined use of standardized defect codes and consistent linking of nonconformances to specific operations, serials, and lots. If defect coding is inconsistent across lines or shifts, the MES will still show numbers but trend and root cause patterns will often be misleading. In practice, many sites need a data cleanup and code rationalization effort before these KPIs support reliable problem-solving.
Cost and schedule impact of scrap and rework
Leadership typically wants to understand not just how much scrap and rework occurs, but their cost and schedule impact. In MES, the most practical KPIs are usually rework hours per unit, rework hours as a percentage of total direct labor, and unplanned rework WIP as a percentage of total WIP. Translating scrap and rework into financial cost per part or program usually requires reconciliation with ERP, which holds material and standard cost data, so these cost KPIs are only as accurate as the integration and cost model. For schedule impact, metrics like rework-related cycle time extensions, number of orders missing planned completion due to open nonconformances, and average time-in-status for rework queues can be tracked if the MES captures timestamps for status changes. In a brownfield stack, you may not be able to get all of this from MES alone; partial automation plus manual analytics is common and should be acknowledged in KPI governance.
Containment and escape KPIs for nonconformances
Aerospace programs often care about how effectively the plant contains nonconforming product and prevents escapes down the value chain. KPIs such as percentage of defects detected at the source operation versus downstream, and number of escapes to subsequent operations or external customers per million units, help quantify the robustness of detection controls. Another useful metric is time to containment for a discovered defect pattern: how long from first recorded nonconformance in MES to defined containment actions implemented on all affected lines. These KPIs require reliable linkage between MES nonconformance records, QMS records, and sometimes customer return data; in many plants, this linkage is partial, so the metrics must be labeled as approximate. Containment KPIs also depend on accurate configuration of inspection steps and sign-offs in the MES, so any unmodeled manual inspections will weaken the signal.
Process and resource-specific scrap and rework KPIs
To make scrap and rework actionable, many aerospace plants define KPIs at the level of critical processes, resources, or configurations. Examples include scrap rate by special process (heat treat, plating, composite layup), by machine or fixture, and by key material batches or suppliers if lot traceability is available in MES. These metrics highlight systematic issues with specific assets or materials but rely on consistent scanning, serial/lot capture, and equipment identifiers in the MES. In mixed-vendor and legacy environments, routing and resource models are often incomplete, which can cause misattribution of defects to the wrong operation or machine. When configuration is weak, sites sometimes start with simpler groupings (e.g., by cell or area) before attempting machine-level KPIs. Any time routing changes or equipment is re-assigned, change control and configuration management need to ensure KPIs remain interpretable over multi-year horizons.
Rework process health: backlog, aging, and closure discipline
Rework itself needs to be monitored so it does not become an uncontrolled parallel process. Useful KPIs here include total rework backlog (units or orders in rework status), aging of rework items (how long units sit waiting for rework), and rework closure rate versus creation rate over time. These metrics pressure-test whether the plant is merely accumulating rework or actually solving underlying problems and clearing the backlog. They depend on having explicit rework states or operations modeled in MES, rather than informal side processes that bypass the system. If technicians perform rework off-record or use generic rework records not tied to specific serials, the KPIs will understate the true problem. In regulated environments, ensuring that all rework paths are represented in validated routings is often as important as the numbers themselves, because unmodeled rework can undermine traceability and conformity evidence.
Linking MES KPIs to QMS, ERP, and root cause analysis
MES scrap and rework KPIs are most valuable when they drive disciplined problem-solving, not just reporting. Plants commonly track the proportion of repeated defects (same code, operation, and part) after a corrective action is closed, and the time from first defect signal in MES to formal corrective action initiation in the QMS. Another useful KPI is the share of scrap and rework volume covered by active corrective actions, which reveals whether efforts are focused on the largest contributors. These metrics require stable defect coding, reliable cross-references between MES nonconformances and QMS records, and clear ownership for investigating patterns. In brownfield stacks, this linkage is often built gradually, using exports and manual joins at first, then more formal integrations once patterns and definitions have stabilized and can justify the validation burden.
Practical constraints and tradeoffs when implementing these KPIs
Not every plant can or should implement every KPI; the practical set depends on the maturity of MES configuration, data collection discipline, and integration with QMS and ERP. Trying to stand up a large KPI catalog on weak data usually produces untrusted dashboards, which erode operator and engineer confidence. Many aerospace sites do better by starting with a minimal viable set (e.g., FPY, scrap rate, defect rate by code, and rework hours) and only adding new KPIs once underlying data gaps and process issues are addressed. Because equipment lifecycles are long and system changes require revalidation, changing KPI logic frequently is risky; it is better to invest time in getting definitions aligned across functions before implementing. Ultimately, the most important MES KPIs for scrap and rework are the ones that can be traced, reconciled, and repeatedly used in day-to-day decisions, even if that means accepting less granularity or slower rollout in the early phases.