I. INTRODUCTIONLECTRICAL machines are essentially wound components for modern industrial systems. During operation, the insulation materials in electrical machines are exposed to thermal stresses. These thermal stresses initiate progressive degradation of the insulation materials, which lead to the deterioration of the insulation characteristics and consequently to insulation breakdown [1]. Therefore, knowledge relating to the condition of the insulation systems and an estimation of the remaining life has become an important concern to manufacturers and users of electrical systems [2]. Real-Time lifetime prediction of insulation materials is of interest in this work. Severe winding insulation breakdown due to thermal stresses have been reported as contributing to as much as 30-40% of all stator winding failures in conventional induction machines [3]. It was also reported in [4][5][6] that thermal ageing is the dominant ageing factor and can contribute up to ≈31% of the deterioration of the insulation materials.The lifespan of the insulation is usually predicted using accelerated ageing tests. An accelerated ageing test is used in the effort to obtain failure statistics in shorter timescales. In [2,7], the lifetime of the insulation material was estimated using an electro-thermal ageing model that takes into account the combined electrical and thermal stresses in the lifetime prediction. Real-time insulation lifetime predictions during steady-state and transient operating conditions of wound components have received less attention in the available literature.The thermal ageing model based on the classical Arrhenius relationship can be adapted for real-time steadystate and transient thermal ageing predictions of the insulation materials for wound components. The key input to the insulation lifetime model is the measured temperature distribution around the winding insulation.Winding temperatures are conventionally sensed using thermocouples or resistance temperature detectors embedded on the exposed external surfaces of the electrical machine and/or windings [8]. These techniques commonly provide single-point temperature measurements that are often unreliable [8]. Moreover, these sensors come with significant challenges imposed by the location of the sensors close to or within the stator windings for example due to use of electrically conductive materials in the sensor packages as well as susceptibility to electromagnetic interference [8]. The key objective in this work is to access the capability of predicting real-time remaining life of wound components insulation materials using Fibre Bragg Grating (FBG) sensors embedded directly into the stator coils during fabrication. The FBG sensors are fibre-optic sensors that have been used in other applications, ranging from frame vibration sensing of the stator core and coil surface thermal monitoring [9][10][11][12]. The FBG sensors are small and can be laid into the middle of the coils during the wind process and can provide real-time temperature information at several...