To increase facility safety by integrating autonomous gas inspection drones with Convolutional Neural Networks (CNNs) and Internet of Things (loT) technologies. Due to their ability to reach difficult situations, unmanned aerial vehicles are increasingly used for gas inspection in industry. However, drone safety and gas detection accuracy remain major issues. It offers a comprehensive method that uses CNN s for real-time image analysis and classification to help the drone locate gas leaks with high accuracy. CNN is trained on a broad dataset of gas leak situations and environmental conditions. loT technology also allows the drone and centralized monitoring system to communicate in real-time for data sharing and decision-making. Our method can adjust to changing lighting and weather and discriminate between innocuous abnormalities and gas leaks. Advanced navigation algorithms let the drone traverse complicated industrial sites while avoiding obstacles and maximizing coverage. Experimental findings show the suggested technology improves facility safety by lowering gas leak reaction times and operator risk. CNNs with loT technologies increase gas leak detection and allow proactive maintenance, improving industrial facility safety and efficiency. It advances autonomous systems for industrial applications by tackling gas inspection safety problems. The suggested framework lays the groundwork for integrating innovative technology to improve autonomous gas inspection drone capabilities and safety. © 2024 Elsevier B.V., All rights reserved.