Motion planning is an important factor affecting the efficiency of industrial robot manipulators. This paper develops an asymmetric 7-segment S-curve motion planning architecture for enabling industrial robot manipulators to complete tasks in the shortest time possible while still satisfying physical constraints. If S-curve motion planning is performed in the Cartesian space, the moving path of the industrial robot manipulator can be specified. However, because of the nonlinear mapping between the Cartesian space and the joint space, physical constraints in the joint space may be violated. To cope with the aforementioned problem, this paper proposes an approach that exploits evolutionary computation algorithms to determine the parameters of the asymmetric 7-segment S-curve so as to achieve time-optimized motion planning without violating the physical constraints in both Cartesian space and joint space. The results of simulations and experiments conducted on a 6-DOF industrial manipulator indicate that the proposed approach is feasible.