Permanent Magnet (PM) machines have attracted much attention due to their high powertorque density, high efficiency, high dynamic response performance, high power factor, and simple structure. Even though there are some uncertainties about their reliable chain supply, PM machines still are the first choice for high-performance applications. Therefore, paying attention to the development and improvement of the performance of these types of electric machines is of great importance. One of the challenges of the PM machines is their vulnerability to Irreversible Demagnetization (ID), especially for applications such as Electric Vehicles (EV), in which the working temperature and Magnetic Motive Force (MMF) produced by stator winding are higher than in conventional applications. Therefore, the ID phenomenon must be regarded as a leading design and control parameter for PM machine designers. The ID can be defined as an irreversible reduction of the remnant flux density of the PM or the EMF of the PM machine due to severe temperature conditions, high external Demagnetization Field (DF), or non-proper design of PM machines, resulting in degradation in performance of the PM machines. The lack of a good source that gives a deep inside into this phenomenon in designing and developing PM machines encourages the author to conduct a comprehensive literature review. For proper design of the PM machines in terms of ID, it is necessary to consider four critical steps, i.e.: (a) accurate non-linear modelling of the PM demagnetization curve, (b) analytical model taking into account local ID, (c) geometry optimization methods targeting ID, (d) accurate experimental testing to evaluate the ID fault-tolerant capability of the manufactured PM machines. In this paper, the above four aspects are surveyed exclusively, and the advantages and disadvantages of the existing methods in these areas are highlighted. Furthermore, design tips towards the design of ID fault-tolerant PM machines and other research opportunities to enhance the reliability of PMs machines are also presented.