HSE ANALYTICS

M7 Data | Business Intelligence & HSE | Marcello

HSE Analytics

"Transforming indicators into prevention and decision-making."

My career as an Occupational Safety Technician and Manager has allowed me to experience firsthand the challenges of risk management in complex environments. Today, I combine this practical experience with Data Science to create solutions that go beyond simple visualization: I translate compliance and incident metrics into preventive strategies. At HSE Analytics, the focus is on intelligence applied to preserving life and ensuring operational continuity, guaranteeing that each dashboard is an active tool in the Safety and Health culture (NR-37 and related standards).

HSE Risk

Problem

Managing workplace safety in industrial operations requires continuous monitoring of incidents, operational behavior, and compliance with safety standards. The central challenge was to consolidate scattered data on accidents, near misses, PPE use, work schedules, and human factors into a single strategic view.

The lack of an integrated analysis made it difficult to identify critical areas, correlate fatigue, overtime, and incidents, and anticipate operational risks. As a result, decision-making was reactive, based only on events that had already occurred, without predictive capacity for accident prevention.

Solution

I developed a Business Intelligence solution focused on predictive analytics and operational risk prevention (HSE), structured as follows:

Data Engineering (Excel): Data processing and enrichment (ETL), including behavioral variables such as fatigue, length of service, use of PPE, overtime, and near-miss records, ensuring a robust and reliable analytical foundation;

Analytical Modeling (Power BI): Construction of strategic indicators using DAX, including Operational Risk Score, HSE Culture Index, Severity Rate, and correlational analyses between human and operational variables;

Predictive Analytics and Simulation: Implementation of trend analysis (forecast) and scenario simulation (what-if), allowing for the evaluation of the impact of actions such as increased training, reduced overtime, and improved use of PPE.

Strategic Design (UX/UI): Dashboard designed with a focus on quick decision-making, using a modern layout, clear visual hierarchy, and highlighting critical alerts and automated recommendations.

Key Analytics

The analysis revealed critical patterns and relevant opportunities for reducing operational risks:

Operational Risk Management: Identification of sectors and shifts with greater criticality, with emphasis on night operations associated with high levels of fatigue and a significant increase in the number of incidents;

Root Cause of Accidents: Mapping the main causative factors, highlighting a strong correlation between human error, low adherence to the use of PPE, and deficiencies in training;

Behavioral Analysis: Identification of a higher incidence of accidents among employees with less seniority and a greater workload of overtime, reinforcing the direct impact of human factors on safety;

Safety Culture (HSE): Creating a security maturity index, allowing for the assessment of adherence to best practices, such as participation in training and recording near misses.

Prediction and Prevention: Predictive analysis indicated a trend of increased incidents in scenarios with high operational load, enabling proactive actions to mitigate risks.

Tool used

Power BI: Data modeling, advanced DAX calculations, interactive visualization, and predictive analytics;

Excel: Processamento, tratamento e enriquecimento dos dados (ETL);

Figma: Dashboard prototyping and design with a focus on UX/UI and analytical storytelling.

Data source

Simulated data based on real-world industrial operations scenarios and occupational safety and health (HSE) indicators, including operational, behavioral, and compliance variables.

Case NR-37

GitHub

Problem

In offshore operations, risk management is critical, and the volume of raw data often masks dangerous patterns. The central challenge was transforming scattered incident records into a predictive intelligence unit. The lack of centralization prevented the identification of correlations between shifts and sectors, hindering efficient preventive actions and the swift compliance with NR-37 (Brazilian Regulatory Standard 37).

Solution

I implemented an end-to-end Business Intelligence solution, structured as follows:

Data Engineering (Excel): Structuring and cleaning (ETL) incident databases and training records to ensure the integrity of KPIs;

Analytical Modeling (Power BI): Creation of DAX measures for monitoring historical trends (2024-2026) and identifying operational seasonality;

Strategic Design (Figma): I designed the layout focusing on high readability and UX/UI, using an alert color palette (Orange & Black) optimized for quick decision-making in high-pressure environments.

Key Analytics

The dashboard provided immediate transparency into the operational health of the unit:

Criticality Mapping: Precise identification of the Drilling sector (27 occurrences) and the Night shift (55 incidents) as priorities for immediate intervention;

Injury Diagnosis: The analysis revealed that Cuts (34%) and Falls (26.8%) were the biggest causes of absence from work, prompting an immediate review of PPE and work procedures;

Proactive Culture: The trend projection for the beginning of 2026 allowed for the anticipation of behavioral audits, transforming reactive data into an anticipatory prevention strategy.

Tool used

Power BI: Data modeling, DAX calculations, and interactive visualization;

Excel: Data processing and database cleansing (ETL);

Figma: Interface design (UX/UI) and Visual Storytelling.

Data source

Simulated data based on real-world offshore operations metrics and historical workplace safety records from the energy sector.

HSE Insights

Risks and PPE

Problem

In offshore production units, safety is the most critical operational pillar. The challenge of this project was to consolidate a massive volume of accident records (more than 5,000 occurrences) to identify the correlation between the use of PPE and the severity of injuries. The lack of a clear view of seasonality and financial impact (days off work) hindered the implementation of preventive risk management and the efficient planning of training programs.

Solution

I developed an operational intelligence tool using Microsoft Excel, focused on transforming raw data into decision indicators:

Data Engine: Structuring the database using advanced formulas and pivot tables for processing large volumes of data;

Vulnerability Analysis: Implementation of compliance calculations revealed that 37.5% of incidents were linked to the lack of proper use of PPE;

Alert Visualization (UX Design): Creation of a high-contrast Dark Mode interface, using a Gauge (speedometer) graph to indicate the real-time risk level of the operation (currently at a critical state of 65.4%).

Result

In-depth data analysis yielded actionable insights for security management:

Severity Pattern: Identification that incidents during the Day shift and in the Production sector (1,803 occurrences) concentrate the greatest risks;

Business Impact (LTI): Measurement of total days of absence, highlighting that events such as "Explosion" generated a loss of productivity of 499 days, justifying investments in occupational health;

Critical Seasonality: Identification of a significant increase in occurrences in the last quarter of the year (reaching 1,660 in December), allowing for the advance planning of awareness campaigns.

Tool used

Advanced Excel: Pivot tables, lookup/reference formulas, and data modeling;

Data Visualization: Speedometer (Gauge) charts, comparison bars, and data segmentation;

Dashboard Design: Applying UX principles and a color-coded traffic light system for immediate readability.

Data source

A simulated database with realistic characteristics of an offshore production unit, built for the purpose of studying occupational safety and health (HSE) indicators.