Challenge
Ensuring strict compliance with Personal Protective Equipment (PPE) requirements is a critical aspect of workplace safety in high-risk industries. Engie, a global leader in energy and infrastructure, is committed to improving worker safety and operational efficiency across its sites. Traditional PPE monitoring relies on manual on-site inspections, which pose several challenges:
- Inconsistent monitoring: inspections happened at fixed intervals, leaving gaps in real-time compliance tracking
- Human errors: subjective assessments may lead to overlooked safety validations
- Scaling limitations: large worksites and multiple locations make it difficult to enforce PPE adherence effectively
To enhance workplace safety and compliance, Engie needed a system to continuously monitoring its sites.
Solution
Together with Engie, we developed an computer vision-based PPE adherence monitoring system that analyses camera footage to detect whether workers are wearing required safety gear, such as helmets and gloves. The system provides real-time analytics and compliance reports, helping site managers enforce safety regulations more effectively.
Key features of the solution include:
- Computer vision-based PPE detection: the model identifies workers and checks for the presence of PPE, with a focus on helmets as a primary requirement
- Integration with existing camera infrastructure: the system processes video feeds from Engie’s on-site security cameras, eliminating the need for additional hardware
- Automated compliance statistics: generate reports on PPE adherence rates, highlighting non-compliance trends
- Dashboard for safety teams: a user-friendly interface enables real-time monitoring and post-event analysis
This AI-powered approach allows Engie to proactively monitor PPE compliance, reducing manual inspection workload while improving workplace safety.
Approach
We followed a structured methodology to ensure a reliable and scalable AI solution:
- Data collection & model training: the team used video recordings from Engie’s worksites to train an AI model capable of detecting helmets and other PPE
- Fine-tuning & accuracy testing: the model was evaluated on real-world site conditions, with adjustments made to minimise false positives and false negatives
- Real-time video analysis: the system was optimised to process live camera feeds and extract compliance statistics
- Dashboard development: a reporting dashboard was created to visualse PPE adherence data, allowing safety teams to track compliance over time
- Iterative refinement: based on initial deployment feedback, the system was adjusted to ensure optimal detection performance
By integrating AI-based video analysis, Engie can continuously monitor PPE adherence with greater accuracy and efficiency.
Outcome
The AI-powered PPE monitoring system provides valuable insights into workplace safety, delivering key benefits:
- Improved compliance tracking: automates PPE monitoring, reducing reliance on manual inspections
- Data-driven safety improvements: identifies compliance trends, helping site managers take proactive measures
- Scalability across worksites: the system can be deployed at large construction sites, power plants, and industrial facilities
- Objective, consistent monitoring: removes human bias from safety assessments, ensuring standardised enforcement
By leveraging AI to enhance workplace safety monitoring, Engie is taking a proactive, data-driven approach to improving on-site compliance, reducing risk, and ensuring worker safety at scale.