In recent years, synchronous reluctance (SyR) machine has been receiving an increasing interest, mainly due to the lack of permanent magnets (PMs) in the rotor, the good torque density, structural robustness and the low cost. The main drawbacks of such a machine are the torque ripple and the low power factor. Nevertheless, they can be efficiently improved with a proper optimization of the rotor flux-barriers geometry. This aspect has been deeply studied in the past and still today, it is the subject of study of many authors. The first part of this thesis continues such a mission investigating the design of SyR machine for challenging fields such as high-speed and fault-tolerant. In high-speed applications both mechanical and magnetic design have to be taken into account to maximize the torque density and, at the same time, to ensure the structural integrity against the high centrifugal forces that arise in the rotor parts. The size of the radial iron bridges play a key role in this contest unlike the tangential ones that can be neglected. In fact, from the magnetic point of view, they should be designed as small as possible to limit the flux linkage that passes through them and to increase the rotor magnetic anisotropy as well. On the contrary, to counteract the centrifugal mechanical stresses on the rotor iron parts, the axial cross section of the radial iron bridges have to be designed accurately to avoid plastic rotor deformations or breakages. A simple magnetic analytical model and a more complex reluctance network of the SyR machine that takes into account these two aspects are described in detail and the maximum power limit curve is computed as well. Some finite element optimizations are carried out to provide SyR motor geometries suitable for different speed ranges. A motor performance comparison between a SyR and surface mounted permanent magnet motor for a high-speed application is made providing interesting results in terms of efficiency and total cost. The fault-tolerant capability of the SyR motor is an intrinsic characteristic due to the absence of rotor PMs. In fact, there is no PM back electromotive force and then no short circuit current or braking torque in case of fault. These features make the SyR motor a promising alternative to the PM synchronous machines. A dual three-phase winding is adopted to further increase the fault–tolerance of the motor drive since it represents a cheaper alternative to the fully redundant and multiphase systems and . The main aim of the study is to find an optimal design solution that allow to achieve good performance in terms of torque density, torque ripple, mutual magnetic coupling and unbalanced radial force on the bearings both in nominal and faulty condition. To do that, the effect of different dual-three phase winding arrangements on the faulty performance are firstly analyzed by means of finite element analysis. Based on these results, two finite element based multi-objectives optimizations are carried out on a 36-slot and 48-slot geometries setting the average torque and torque ripple, in healthy and faulty conditions, as costs functions. The most promising design solutions provided by the optimizations are manufactured and their fault-tolerant capability is evaluated experimentally confirming the reliability of the proposed design solutions. The second part of the thesis deals with the development of bonded magnet mixtures to improve the torque density and power factor of SyR machines. In particular, the magnetic mixtures are made with different percentage of neodymium iron boron magnetic powders and binder. Some mixtures are prepared and characterized in laboratory to get the main magnetic properties such as coercive field, remanence flux and the relative magnetic permeability. The performance of a SyR motor, designed and manufactured for the purpose, are evaluated with the different bonded magnets by means of finite element analysis and experimental tests. Good results in terms of torque density and power factor are obtained confirming the bonded magnets suitability for SyR machine improvements. The design and experimental validation of a PM assisted SyR (PMASyR) machine for a low-voltage electric scooter is described in the last part of the thesis. The multi-objective design of such a machine has been deeply investigated in literature. In these works, the average torque, the torque ripple and the constant power speed range are set as cost functions and only the transversal motor geometry is optimized. Furthermore, the number of winding turns and the motor stack length are properly adapted after the optimization to meet the rated Volt-Ampere ratings. However, this is possible only if the number of turns is high enough that, a relatively small modification of the number of turns or stack length, does not involve significant changes on the motor performance. Conversely, if the voltage limit and the maximum target speed yield a very low number of turns any modification of the number of turns or the stack length may involve completely different machine inputs or, more drastically, compromise the feasibility of the drive system. The novelty of the study is to propose a multi-objective PMASyR motor design procedure that (i) evaluates the feasibility of the drive system in overload and flux-weakening condition and (ii) computes the optimization cost functions in the same operating condition to allow easy and direct considerations of the Pareto fronts, without any further computations and checks after the optimization. A prototype of the most promising design solution is manufactured and tested to prove the effectiveness of the procedure.