Abstract: Background　The casualty evacuation equipment currently used by our military mainly includes field ambulances and field stretchers. In modern warfare, it is common to see large numbers of casualties requiring urgent evacuation and treatment within short periods of time. However, current casualty evacuation methods struggle to cope with such demands. Objective　To design and develop a kind of intelligent transport device for the wounded on battlefield to realize unmanned transport and solve the problems of traditional transport device (field ambulance, field stretcher) such as difficult timely casualty transport, low transport efficiency and high casualty risk of rescue personnel on battlefield. Methods　This study adopted a multi-center collaborative R&D model. In system construction, the intelligent casualty transport device integrated the autonomous driving platform with the medical module. In the performance verification design, various vehicle parameters and adaptability of the intelligent transport vehicle in both fully autonomous driving mode and manual remote control mode were tested on two types of terrains (road and off-road). Fifty repeated casualty transport simulation experiments (with a standard distance of 2 kilometers) were set up to compare the fully autonomous driving mode with the manual remote control mode. The main evaluation indicators included transport time, trajectory deviation, speed control accuracy and system failure rate. Results　The length of the intelligent transfer vehicle was 2400 mm, with width of 1 400 mm, height of 2230 mm and total weight of 850 kg. Its maximum load was 500 kg, the maximum electric mileage was 100 km (off-road), and the maximum climb was 32 degree. The test results of the intelligent transport vehicle on both road and off-road conditions showed that the obstacle perception distance in the autonomous driving mode was significantly longer than that in the manual remote control driving mode (30.80 ± 1.59 m vs 29.75 ± 1.28 m, P=0.03 road; (29.00 ± 2.27) m vs 30.78 ± 2.41 m, P=0.02 off-road) and the trajectory deviation in the autonomous driving mode was significantly smaller than that in the manual remote control driving mode (0.70 0.61 - 0.88 vs 0.82 0.76 - 1.00, P=0.03 road; 1.00 0.96 - 1.10 vs 1.15 1.00 - 1.39, P=0.02 off-road). There were no significant differences in other test parameters between the two driving modes on both road and off-road conditions. The median transport time of the intelligent transport vehicle was significantly shorter than that of the manual remote control driving transport vehicle (8.47 min vs 13.26 min, P<0.001). There was no significant difference in the vehicle failure rate between the two driving modes (4% vs 0, P=0.50). The single-point horizontal positioning accuracy of the intelligent transport vehicle was 1.51 m. The lateral deviation during driving was always less than 0.5 m and the longitudinal deviation was less than 0.5 m. The positioning accuracy of the injured was 5.5 m. The intelligent transport vehicle had good obstacle avoidance function and one click return function. Conclusion　This intelligent transport system significantly enhances the efficiency and safety of transporting wounded soldiers on the battlefield. Its modular design and pure electric drive provide key technical support for the construction of a modern combat rescue system.