Variable speed induction motors controlled by variable frequency drives are used for a variety of industrial applications. Monitoring and prognostic occurrence of faults in induction motors is vital for reducing the downtime and accidents. The proposed work focuses on failures in induction motors owing to bearing misalignment and insulation failure in the stator that results in abnormal vibration and temperature rise in the motor. This research intends to improve the dependability and safety of industrial operations by identifying faults in their early stages using advanced methods such as vibration analysis and thermal monitoring. This work focuses on fault prognosis in induction motor through vibration data, which is analyzed using Daubechies orthogonal db10 wavelet transformation. The neural network algorithm optimizes the analyzed results to enable real time fault detection. The temperature of the stator is measured to estimate the expected lifetime of the insulator. The real time vibration and temperature data is measured and transferred to prognostic model build in MATLAB using ATMEGA 32 controller and the results are validated for good, allowable and not permissible conditions of motor based on ISO 10816 vibration levels for Class I motors. The improved accuracy and efficiency of real-time fault detection have the potential to reshape maintenance strategies and enhance the overall reliability of variable speed induction motors.