Background: In the paper the chosen speed and current sensor faults detectors for vector controlled induction motor drive system are presented and described. Systems based on the artificial intelligence (neural network) and simple algorithmic systems are analyzed and tested in different drive conditions.The compensation strategy was proposed and tested. A fault tolerant drive, based on hardware redundancy, was developed and presented. Material and methods: Simulation and experimental results are obtained in direct field oriented control algorithm (DFOC) on the laboratory set-up with rapid prototyping card Micro Lab Box DS1202 by dSpace. Results: Proposed solution is based on three neural networks with three hidden layers and different input signals. The compensation algorithm of faulted sensors based on drive system control topology changes is described. Moreover, in the paper, the results of the fault-tolerant motor drive operation during different fault scenarios are presented. Conclusions: It was proved that proposed complete fault-tolerant algorithm is able to identify faulted sensor and to compensate failure providing further stable operation of the drive. Moreover, presented system is able to ensure the continuation of the process even in case of all measurement sensors failure occurrence in short period of time.