Comprehensive video analytics at the Mahalaxmi Group plant
The Mahalaxmi Group coking plant faced a number of production and organizational issues.
Project background
The Mahalaxmi Group coking plant faced a number of production and organizational issues. Coke oven charging operations were performed with deviations, leading to material spillage and reduced efficiency. Compliance with occupational safety requirements (use of PPE) and labor discipline (use of phones, inactivity) was monitored visually, which did not ensure full coverage. There was no system for automatic tracking of freight vehicle entry/exit, access control to the territory, or monitoring of equipment movement in hazardous areas.
A comprehensive automated solution was required for monitoring coke oven charging, industrial safety, access and logistics management, as well as staff working time tracking.
Implementation progress
Specialists from OSMOVIEW LLC developed and implemented a comprehensive video analytics solution based on the OSMOCV platform. The platform was deployed on the plant's server infrastructure within a local network.
Video streams from IP cameras (including PTZ cameras for dynamic tracking) installed in key production areas were integrated into the platform: charging ovens, perimeter, entry gates, and loading/unloading zones.
The solution included the following in-house neural services developed by OSMOVIEW LLC:
- “Oven Charging Control” — detection of the charging machine position, analysis of positioning accuracy, and identification of deviations;
- “PPE Detection” — recognition of helmets, safety glasses, and masks worn by personnel;
- “Presence/Absence in Zone” — analysis of employee working time in key zones;
- “Mobile Phone Detection” — logging of phone usage in prohibited zones;
- “Movement on Platform/Hazardous Zone Detection” — detection of people on moving platforms and in hazardous zones;
- “License Plate Recognition (LPR)” — automatic logging of truck entry/exit;
- “Fire Detection” — early detection of fires;
- “Digital Tag Recognition” — tracking of equipment and object movement;
- “Face Recognition (FRS)” — identification of employees at the gate for working time tracking and access control.
Additionally, business logic with geofences for access control was configured, along with the generation of alarm events upon violations (incorrect positioning, lack of PPE, presence in a hazardous zone, phone usage) and automatic sending of notifications to operators and managers via Telegram and email. System calibration and testing were conducted, and the client's personnel were trained on how to use the platform.
Achieved results
- Improved oven charging efficiency — the system ensures control of charging machine positioning accuracy, which allowed eliminating material spillage and increasing productivity.
- Reduced number of industrial safety violations — automatic detection of PPE and presence in hazardous zones reduced the number of violations and created a deterrent effect.
- Automated working time tracking — face recognition at the gate eliminated the need for manual timesheets and improved tracking accuracy.
- Strengthened access control — geofencing and object movement detection ensured prevention of unauthorized access and theft.
- Optimized logistics — automatic recognition of truck license plates and integration with the invoice system reduced trip processing time and minimized the risk of errors.
- Ensured 24/7 monitoring — the system operates 24/7, providing continuous control of all production processes.