This work is based on the digital twin intelligent mine explorer, with the digital twin model platform to receive the data collected by the robot, and display the mine model, at the same time through the VR wearable device remote control robot. This work will be divided into three modules to explain. The first part is the bottom module of the car, which is controlled by the STM32F4 development board and JESTON nano development board to control the driving, radar navigation and visual recognition of the robot. The second part is the car detection module, which is controlled by the STM32F4 development board and can detect the environment under the mine. The third part is the digital twin model platform, which receives robot data through the server, use UE4 for VR control, modeling viewing, and more. The bottom module of the first part of the car, the main control is a STM32F4 development board, the peripheral ADC on the board is used to read the battery power, and the other uses the FreeRTOS system, which can manage the gyroscope and the motor at the same time. The gyroscope model is MPU9250, which is used to read the four elements of robot attitude data and send this data to the STM32F4 for subsequent kinematics calculation. The robot is a tracked vehicle with photoelectric encoder and DC deceleration motor drive. At the same time, the STM32F4 development board is connected with a JESTON nano development board, and the inertial measurement unit (IMU) information is transmitted to JESTON nano, and the robot operating system ros is used for kinematics calculation. JESTON nano uses ros to connect an HDMI display, a radar and a camera. The radar scans the panorama under the mine, generates a two-dimensional grid map, regenerates a global cost map and a local cost map, and uses the line-related algorithm to navigate the line and avoid obstacles under the mine. The camera plus a searchlight to fill the light, obtain the underground image data, use the yolo neural network algorithm, identify the image of the underground staff captured by the camera, determine whether they are wearing a safety helmet, and make a corresponding reminder. Part II The STM32F4 development board sends the collected data to a serial port screen, and then displays the data in real time. The above data is also sent to the ESP32 module, and then the data is sent to the server through the TCP protocol for the operation of the DT platform. In the third part of the digital twin model platform, the server sends the received data to the MySQL database, and displays the data collected by the robot in real time through the UI interface of UE4 software. The mine model built by radar is also displayed in UE4, which can be viewed from the third person view or the first person view of the robot. The webcam mounted on the vehicle is retrieved for more sophisticated analysis of downhole data. Wearing VR equipment, with the help of SteamVR, you can also see underground modeling through VR glasses, and use VR handles to remotely control the movement of the robot.