Abstract: As a high-risk industry, offshore oil development requires the establishment of a quantitative assessment system for safety governance. Taking China Oilfield Services Limited (COSL) as the research subject, this study innovatively proposes a "dual-dimensional safety indicator system" (generic indicators + specific indicators) based on Heinrich's Law and the Swiss Cheese Model theoretical framework. It has developed a risk early-warning system covering the entire operational lifecycle. By introducing 12 quantitative indicators such as Crew Stability Rate and Risk Conversion Efficiency Coefficient, the risk perception accuracy has been improved by 47%. Empirical studies demonstrate that this system achieved a 62% reduction in recurring hazard incidence and maintained the OSHA recordable incident rate 35% below the industry average, establishing a closed-loop management paradigm of "indicator-driven process, cultural permeation, and behavioral correction". The research outcomes provide methodological reference for safety governance in highly complex industrial scenarios.

Key words: safety indicator system, risk early-warning model, risk control and management, work safety situation assessment, safety intervention culture