Abstract: A self-organizing artificial social position method is proposed for forming complex swarm-robot transport formations. For each robot R_i at each sampling time, the approach works as follows. Firstly, R_i's handling-point is determined by an artificial social position method. Then, R_i's artificial moment motion controller makes it move one step towards its handling-point. Because the method for attractive segments and attractive points in the controller is improved, the computational burden is decreased, and the controller can still make each robot move safely and finally reach its handling-point. The above steps are repeated until the desired formation is achieved. This method is easy to understand and its computational burden is low. Furthermore, no matter how complex the transport formation, the desired formation can be always achieved by self-organization. Simulations indicate that the proposed method is feasible and efficient.