Abstract: Data-driven detection technology is an effective part of development of smart coal mine. It can predict behavior of future object system without knowing accurate analytical model of smart coal mine big data system. For the difficult problems of intelligent production operation, inherent safety production,scientific operation mode of smart coal mine, application status of data-driven detection technology in fault diagnosis of coal mine equipment, detection of foreign object in belt transportation and detection and identification of coal-gangue was analyzed.The development trend of data-driven detection technology in these three aspects was prospected: ① Fuzzy mathematics and artificial neural network should be integrated into fault diagnosis of coal mine equipment more effectively; ② Video anti-shake, image segmentation and target detection technologies should be more effectively integrated into detection of foreign object in belt transportation; ③ Sorting robot, computer vision and image recognition technologies should be more effectively integrated into detection and identification of coal-gangue, and improving universality of the algorithm would be one of the development directions of coal-gangue image recognition.