"Learning from Mistakes" or How to Conduct a High-Quality Incident Investigation

From Blame-Seeking to Systemic Analysis: Why the A3 Format Matters

In modern production culture, incident investigation is not a tool for punishment, but an opportunity to identify systemic failures and prevent their recurrence. Anton Dolgikh, Leading Specialist in HSE Tools Development at OEMK (Metalloinvest), explains why traditional methods often fail to yield the desired results and how transitioning to the A3 format helps find the true causes of incidents.

The speaker emphasizes that 96% of all problems are caused by an improper management system, and only 4% by intentional errors of performers. Punishing employees for every mistake only leads to hiding problems out of fear, demotivating staff, and preserving systemic flaws. Instead of looking for a scapegoat, it is more effective to analyze the root causes of failures and change processes.

Comparison of Analysis Methods: SAP, RCA, and A3 Format

The presentation details three main systems for analyzing the causes of incidents:

• SAP System: Useful for analyzing technical factors during equipment downtime and resource planning. However, it does not allow for a deep determination of systemic root causes, limiting itself to recording symptoms.
• RCA (Root Cause Analysis) Method: Helps work through the details of an incident using a cause tree. Suitable for both downtime and accidents. The main risks are stopping at a superficial cause or blaming people, as well as the lack of verification of the effectiveness of implemented measures.
• A3 Format: The method chosen by Metalloinvest, combining a range of tools for deep analysis. It generates a structured report and is universal for investigating any incidents, requiring the participation of a competent independent team and a methodologist.

Step-by-Step Investigation Algorithm in A3 Format

The speaker demonstrates by example how a high-quality investigation process is built, starting with the categorization of the incident according to a criticality matrix (from minor to catastrophic level). This determines the composition of the commission and the depth of the analysis.

• Scene Inspection: Must be conducted in a timely manner, without narrowing the boundaries of the inspection or jumping to conclusions. Preparing a checklist and recording all facts is important.
• Interviewing Participants and Witnesses: Using a questioning funnel (from general to specific), considering non-verbal communication, and mandatory data recording.
• Document Collection and Event Reconstruction: Analysis of documents on personnel, processes, and equipment. If necessary, safe recreation of the situation to obtain additional data.
• Cause-and-Effect Analysis: Applying a timeline to systematize facts, the "5 Whys" method, and a cause tree to identify root factors. The Ishikawa diagram helps structure these causes by category.
• Development of Corrective Actions: Building a barrier diagram (existing and required) and formulating actions according to the SMART system. Actions are evaluated using an "effectiveness-cost" matrix, prioritizing highly effective and low-cost solutions.

Lessons Learned and Transparency

The final, yet critically important stage is disseminating the investigation results. If a company does not share its findings, it is doomed to repeat its mistakes. Open knowledge sharing through corporate portals, meetings, and information boards prevents repeated failures, saves resources, and enhances the overall safety culture.

What You Will Learn from This Webinar:

• How to properly classify incidents and form an investigation commission?
• Why seeking blame reduces safety levels and how to transition to analyzing systemic causes?
• Which tools (timeline, cause tree, Ishikawa diagram, barrier diagram) to use for deep incident analysis?
• How to evaluate and prioritize corrective actions using the "effectiveness-cost" matrix?
• How to build a system for sharing lessons learned to prevent similar incidents in the future?