Three-Level Indicator System for Engineering Quality Management

Peter Drucker said “No data, no management.” This article explains a three‑level indicator system used to quantify engineering quality, turning qualitative descriptions into measurable metrics.

Why three levels? Level‑1 (result) indicators give a high‑level view but are lagging and coarse. Level‑2 breaks them down for targeted improvement, and Level‑3 provides actionable improvement metrics, enabling drill‑down analysis and closed‑loop improvement.

The system focuses on four dimensions: efficiency, quality, stability and resource utilization.

Efficiency is measured by throughput, demand‑to‑delivery ratios, planning capability, and collaboration efficiency (e.g., estimation accuracy, test automation rate, engineering availability).

Quality refers to built‑in product quality. Key metrics include defect introduction rate, defect count and distribution, and gate‑keeping criteria that filter low‑quality artifacts before testing.

Stability captures production incident frequency and severity. Indicators such as P1‑P4 incident counts, incident severity × recovery time, and backlog closure rate are used to assess and improve system stability.

Resource evaluates human resource allocation: headcount, focus areas, and efficiency of effort (e.g., development‑test ratio, pre‑implementation ratio, average test cases executed per hour).

Two practical scenarios are presented:

Drilling down a “development‑test ratio” of 4.1:1 to its constituent level‑2 and level‑3 metrics, revealing demand throughput and resource distribution.

Applying a “red‑green‑light” risk‑warning mechanism that combines quantitative trends from gate‑keeping quality, built‑in quality, and collaboration efficiency with qualitative notes to flag potential quality risks before release.

The conclusion reiterates that quantitative, data‑driven indicators provide a systematic, sustainable asset for R&D digitalization, supporting continuous improvement, risk mitigation, and efficient delivery.