In order to optimize the development of automated logistics, this paper proposes to build a mobile robot based on ROS to carry out intelligent autonomous navigation and distribution in the warehouse environment and to use intelligent robots to improve the efficiency of logistics and distribution. Analyze logistics robots’ localization methods and propose a SLAM algorithm to obtain environmental map information. Improve the Manhattan distance function to propose a global path planning A* algorithm based on the two-way search of the initial point and the target point, as well as a local path planning DWA algorithm to seek the best path. Collecting robot scheduling principles and scheduling conflict types, a genetic algorithm was selected for workshop AGV scheduling planning. Set up an experimental environment to analyze the optimal performance of the DWA algorithm. Combined with the intelligent warehouse goods sorting environment, analyze the planning and scheduling strategy for the goods sorting path of the ROS robot cluster. In the experiment, it is concluded that the larger the value of parameter β of the trajectory evaluation function is, the longer the time to be consumed by the calculation of the DWA algorithm is, and the designed path distance will be farther. When α =0.02 and β =0.2, the DWA algorithm reaches the best. Combined with the actual task scheduling path of the mobile robot cluster, it can be considered to be able to complete the autonomous navigation experiment in the whole warehouse environment, which is in line with the demand for automated logistics use.