Purpose
Plants possess equipment, labor, schedules, standards, maintenance systems, and years of operating data. Yet the production plan still fails because the reported average hides instability, the diagnosis points to the wrong owner, or no routine connects a finding to a verified fix.
This program examines the decision layer between measurement and action.
Current work
Throughput Reliability Index
TRI extends—not replaces—OEE by asking whether current performance is dependable and where instability changes the planning decision. The public method describes the conceptual factors and validation requirements. Detailed calibration remains unpublished until it has sufficient client and cross-site evidence.
Deep Constraints Triage
A public workflow for structuring the first response to a shift-floor problem, identifying immediate safety or escalation boundaries, and producing a cleaner passdown.
Questions under review
- When does a reliability adjustment improve decisions beyond OEE and ordinary SPC?
- How much observed variance belongs to crew, equipment, schedule, material, or the measurement system?
- Which early-warning patterns survive prospective testing rather than only retrospective explanation?
- Can economic exposure be estimated without creating false precision?
- What minimum evidence should be required before an automated recommendation is actionable?
Validation boundary
TRI is a working method, not a validated universal standard. It needs prospective field testing, cross-line comparison, sensitivity analysis, and evidence that it changes decisions or predicts operational risk better than simpler baselines.
Revision history
- v0.1 — August 7, 2026: Established the public industrial research program and separated public method from private calibration.