Abstract: The theoretical and technical study on intelligent roadway development is very important to solve its low efficiency problems in coal mines. The remote intelligent controlling is the fundamental goal and strategic vision of using less people or unmanned operation, facing the bottleneck technical problems such as single driving-gear positioning and control, equipment group coordination, man-machine-surrounding perception and rendering, and remote intelligence decision. Aimed at the requirements on the remote intelligent controlling, this paper proposes a technical framework for remote intelligent control based on DT (Digital Twins). The DT models of intelligent remote driving is built, and the underground personnel, equipment and environmental information are presented in the digital space, realizing the fusion and the mutual intelligent driving of virtual and reality, achieving the intelligent collaborative control of digital driving and the physical driving. So the safety problems in roadway excavation are solved. Aiming at the requirements of the parallel operation of the mining and transport, the remote control framework of automatic drilling anchor and efficient transport is proposed, and the research progress of remote virtual rendering, accurate position perception, DT data sharing, false synchronous drive, process memory cutting and equipment group collision alert are introduced. The paper introduces four technologies such as "DT+VR" remote control decision, "visual +" position estimation, "artificial teaching" memory cutting, and "virtual equipment" collision warning, to overcome the problems in intelligent decision-making, the accurate positioning, the trajectory planning and the equipment group collision warning in underground coal mines. The basic theories and technologies, such as the refraction imaging mechanism modeling and modification of plane flameproof glass and optical ball cover, imaging blur mechanism under vibration condition, the accurate body pose estimation of roadheader, directional navigation and deviation correction, manual teaching cutting track and memory cutting based on visual measurement, devices cooperative control and digital twin driven remote control, are summarized. Based on the imaging refraction correction and de-blurring of visual perception, this paper systematically overviews the progress of vision technology in positioning, orientation and shape cutting, and also analyzes that digital twin drive technology is an effective way to realize the normal production of intelligent remote roadway development. The core technologies mentioned above have been tested in the laboratory and preliminary verified underground, for instance, at the Xiaobadang Coal Mine and the Dahaize Coal Mine in Shaanxi province, China. The researches provide a new realization path for achieving remote intelligent controlling in coal mines.