Abstract: Aiming at the low intelligence level of trackless rubber-tyred vehicles in coal mines, the harsh working environment, insufficient lighting, high water mist, poor road conditions, complex road conditions and networks, and large visual blind spots of drivers, an intelligent assisted driving system based on multi-sensor data fusion is designed. Firstly, the current status and trend of intelligent application of trackless rubber-tyred vehicles are analyzed; secondly, the framework structure of the multimodal data fusion intelligent assisted driving system is designed, including hardware architecture and software architecture; finally, the specific functional design of the intelligent assisted driving system is given. The test results show that the accuracy of the multimodal fusion perception algorithm of the system in obstacle recognition has reached more than 96%, and the safety of man-machine cooperative driving is improved to 99.6%.