Optimization of safety on construction sites with using artificial intelligence to recognize construction equipment

Abstract

The paper explores the application of artificial intelligence (AI) to identify construction equipment using CCTV cameras on construction sites. The main goal is to increase safety levels, optimize construction processes and increase productivity. The research covers the analysis of world and Russian experience in creating neural networks, collecting data on construction equipment and developing an AI model for recognizing and classifying machines based on their visual characteristics. The system was also tested to evaluate its accuracy and efficiency. The article discusses various approaches to training neural networks, including supervised and unsupervised learning, reinforcement learning, transfer learning, generative adversarial networks, convolutional and recurrent neural networks, autoencoders, and hybrid models. Particular attention is paid to integrating AI with other aspects of construction site safety, including alarm systems, emergency response and personnel training. This research proposes a multi-layered approach to construction site safety, ensuring not only accident prevention but also improved work processes. Ultimately, the application of AI on construction sites helps create a safer, more organized and controlled work environment, which is critical for large construction projects.