There are various challenges in designing a life-cycle robust permanent magnet actuator (PMA) to satisfy the application-oriented requirements of quality consistency. In particular, ubiquitous uncertainties from manufacturing and operating process that cause unstable or even infeasible solutions significantly increase the difficulty of the life-cycle robust design of PMA. Conventional methods usually treated these uncertainties as fixed design constraints, which neglect the nonlinear characteristic of time-dependent degradations, cannot reflect authentic impacts on the life-cycle robustness of PMA. To address this issue, this study propose a method of life-cycle dynamic robust design optimization for batch production PMA. An innovative time-dependent global sensitivity analysis-based classification method is first proposed to divide design variables into three categories: nonlinear, degradation insensitive, and degradation sensitive variables. Then, a modified Spatiotemporal Kriging model with Karhunen-Loeve expansion is applied to establish the life-cycle robustness model of PMA with consideration of uncertainties. This model can valid describe the nonlinear fluctuation of the life-cycle robustness. Furthermore, on the basis of the three categories of design variables and life-cycle robustness model, a multistage dynamic robust design optimization strategy is derived, which realizes the synchronous control of life-cycle robustness deterioration rate (mean value) and fluctuation (standard deviation). Finally, taking a case of rotary PMA, the effectiveness of the proposed method is verified.