Because of the inherent dangers of the job, virtual reality (VR) is in great demand for use in coal mine safety and rescue drills. VR has made great strides in the last decade toward better hardware and software for miner training. But there are still some downsides, such as pricey and inappropriate technology, an unsatisfactory user experience, no direct browser access, and an intelligent and humanized design that is missing. This research proposes a cloud-based VR system for coal miners' training as a solution to these issues. The eight modules that make up the system show the whole process of an underground coal mine. It has a browser/client architecture. A better experience for users using browsers is achieved using online cloud-rendered video streaming, which provides sufficient computation and rendering capability. To further enhance the system's ability to communicate with users on an emotional level, gaming artificial intelligence (AI) and Support vector machine (SVM) are also included in the systems. To improve the training experience, this system creates two virtual miners, which is different from the usual VR training software. A non-player character (NPC) with a focus on tasks, the first virtual miner educates players about the mine as a whole and shows them the ropes when they go down to do their job. Users may become ready for common calamities with the help of the second virtual miner, who is disaster oriented. Successful implementation and validation of the system's performance have been carried out in a controlled laboratory setting. However, further procedures are required to encourage the development of novel uses for VR in disaster drilling and miner training.