Sinusoidally fed permanent magnet synchronous motors (PMSM) fulfill the special features required for traction motors to be applied in electric vehicles (EV). Among them, axial flux permanent magnet (AFPM) synchronous motors are especially suited for in-wheel applications. Electric motors used in such applications must meet two main requirements, i.e. high power density and fault tolerance. This paper deals with the optimal design of an AFPM for in-wheel applications used to drive an electrical scooter. The single-objective optimization process carried out in this paper is based on designing the AFPM to obtain an optimized power density while ensuring appropriate fault tolerance requirements. For this purpose a set of analytical equations are applied to obtain the geometrical, electric and mechanical parameters of the optimized AFPM and several design restrictions are applied to ensure fault tolerance capability. The optimization process is based on a genetic algorithm and two more constrained nonlinear optimization algorithms in which the objective function is the power density. Comparisons with available data found in the technical bibliography show the appropriateness of the approach developed in this work.