Strategic impact: safety and scale in heavy industry
The client is one of the world’s largest steel producers. With massive metal loads constantly moved by overhead cranes, operational safety is the company’s central pillar. To raise the bar further, a computer vision solution was developed to protect assets and the physical wellbeing of operators in real time.
The challenge: fatigue in remote control rooms
Modernised control rooms brought ergonomic gains, but also invisible risks linked to the comfort of remote work. The main challenge was to monitor operators’ alertness non-invasively, preventing risks caused by:
- Cognitive fatigue. Silent signs of tiredness during long shifts in quiet environments.
- Reduced alertness. Momentary lapses of attention while moving heavy loads.
- Operational vulnerability. Risk of large-scale accidents and material damage caused by human error.
The solution: multimodal artificial intelligence
An applied-intelligence architecture that analyses behavioural patterns using deep learning algorithms.
Advanced biometric monitoring
High-definition cameras combined with AI models that detect signs of tiredness through several indicators: eye aspect ratio (EAR), yawning frequency (MAR), head orientation and shoulder posture.
Real-time preventive alerts
The system processes images locally and automatically recognises when the combination of indicators signals a critical state of fatigue. Alerts go straight to the security centre, allowing intervention before any risk materialises.
Governance and traceability
Every event is logged on an auditable platform, giving safety management historical data to continuously improve processes and shift patterns.