To achieve automated crawler crane path planning, environment construction, modeling, algorithm design, simulation, and real vehicle testing are studied. For map reconstruction, environmental data is gathered using unmanned aerial vehicles (UVA). And a 3D model is built processed by mapping software, and is converted into 2D gridded height information through bounding box algorithms for path planning. The crane’s mathematical model is built according to forward and inverse kinematics. Then, the connection between the end pose of the lifted object and the joint angles is established. In path planning, the RRT algorithm is enhanced with motion constraints and heuristic strategies, so as to improve the smoothness and search efficiency of the paths. Simulation comparisons show that the improved algorithm can generate better paths than other algorithms. Real testing confirms that the lifted object is moved quickly and stably to the target node, successfully avoiding obstacles.