Engineer-to-order (ETO) and configure-to-order (CTO) companies reuse modules across product variants while continuously maturing them through concurrent engineering processes. In this setting, readiness for module release depends not only on the module’s own state but also on the readiness of every revision it requires. Because modules are shared across variants, a readiness contradiction in one dependency can block release across multiple products. Previous research has addressed variant governance, change propagation analysis, and effectivity-enabled product structures, but has paid little attention to engineer-facing, explainable checks that diagnose present-state readiness contradictions across the product structure. To address this gap, this paper proposes a Revision Coherence Checking method that (i) maps enterprise evidence — lifecycle state, change-task completion, approvals, and effectivity rules — into canonical readiness states via an auditable interpretation layer, and (ii) checks coherence constraints over a product-structure dependency graph to identify violations where a module revision is treated as more ready than a required dependency. The method is operationalized through an interactive graph-based prototype that localizes blockers, exposes propagation paths, and supports prioritization using reuse impact. A case study at a European custom laser manufacturer, covering 24 product lines and 183,265 dependency relationships, identified 532 direct readiness contradictions not surfaced by the company’s existing PLM system. The results demonstrate that the method can detect and characterize coherence violations, enabling earlier detection of release blockers, reduced rework from late discovery of readiness mismatches, and more targeted engineering coordination. Overall, the study frames readiness coherence as a computable, structure-level property of revision-aware product structures, rather than an attribute of individual revisions, thereby providing a basis for diagnostic readiness reasoning in reuse-intensive, PLM-enabled engineering environments.
Kourtis et al. (Fri,) studied this question.