How AI in Construction Helps Protect Workers' Health: A Case Study of a Construction Company
A construction company implemented a custom AI system combining cameras, sensors, and neural networks to enhance workplace safety, automate transport accounting, and control project timelines. The system reduced accidents by 30% and also helped prevent material theft and monitor construction progress against BIM models.
A large construction company approached the developer (RedKrab) with a request to automate safety compliance checks on construction sites, because manual control was insufficient: safety engineers cannot be everywhere, control was reactive, and data did not match reality. The custom system uses video cameras, location sensors, and neural networks. Cameras monitor entrances and exits for vehicle and material accounting, control personal protective equipment (helmets, gloves, etc.), detect unauthorized access to dangerous zones, count staff, and monitor loading operations. Sensors provide staff positioning with accuracy up to five meters, useful for emergency response and geofencing alerts. Video analytics tracks construction pace and compares progress with BIM models. The system sends alerts via SMS and email, stores violation video fragments, and maintains a central incident database. The company reports a 30% reduction in accidents and related costs, though savings from prevented theft or schedule delays are not yet confirmed. The approach is applicable to other distributed workplaces, industrial enterprises, warehouses, and mining operations.
- Abbreviations
- SMS = Short Message Service — служба коротких сообщений
- BIM = Building Information Modeling — информационное моделирование зданий
Source: Habr — хаб ИИ —
original