Path planning, as a primary mission in crane lift planning, has a profound and direct impact on the safe and efficient execution of lift jobs on construction sites. Typically, the main objective of path planning is to find the shortest (or a relatively short) and collision-free path from the load supply point to the demand point in a finite 3D space with the presence of obstacles, while considering the mechanical constraints of the crane. Despite collision-free being a primary criterion in path planning, other critical safety issues have not been adequately addressed in existing methods. For example, blind lifts and reduced visibility of crane operators are prevalent in construction and have been recognized as major safety concerns in crane lifts. Furthermore, complex coordination of crane motions and frequent changes of moving direction potentially lead to human errors. To address this gap in the knowledge, this paper proposes a semi-automated planning approach by using Building Information Modeling (BIM) and an intelligent path finding algorithm. First, dimension and weight data of building components to be lifted are retrieved from the BIM model. Then, a modified RRT* algorithm is used to generate a short and collision-free path that satisfies the desired level of path smoothness, visibility, and motion coordination. Finally, paths for all lifts are stored and associated with corresponding elements in BIM for easy analysis and visualization. Preliminary results show that the proposed method can effectively reduce the occurrence of blind lifts while ensuring a practical path for execution. In the future, the proposed method is expected to enable a BIM-based risk analysis tool for the safety of crane operation and its impact on other adjacent construction activities.