In this paper to deal with the sensor faults, Non-singular Terminal Sliding Mode Fault Tolerant Control (NTSM-FTC) is proposed. In comparison to existing sliding mode control (SMC), the proposed controller can retain the advantages of the SMC and it can also improve the system performance and make the systems states converge to equilibrium point in a finite time. Moreover, it can solve the singularity problem in terminal SMC method. This controller utilizes a Fault Detection and Diagnosis (FDD) system based on bank of Adaptive Extended Kalman Filter (AEKF). Conventional Extended Kalman Filter (EKF) suffers from performance depreciation and may even diverge in case the statistics of the noises were unknown. Hence, in this paper AEKF has been proposed to maintain the desired performance, while the statistics of the noises are unknown. Next, by simulating sensor faults on a two degree of freedom robot manipulator, the capability of the proposed FTC in terms of stability ensuring in comparison to conventional SMC has been shown and the superiority of the proposed AEKF in terms of estimation and FDD accuracy has been illustrated.