Abstarct: Autonomous systems, where the human operator is frequently removed from the loop, are more dependent on the increasing numbers of sensors to acquire system information. This, in turn, makes systems more vulnerable to faults in sensors. In this thesis, sensor fault detection for a class of nonlinear systems is investigated. Fault taken into account here is the collective form. Fault taken into account here is the collective form. Then, using a modified structure and characteristics of the filter system designed sliding mode observer to reconstruct a sensor fault in the system is provided with uncertainty. In this work an adaptive sliding mode observer for nonlinear systems with parametric uncertainty fault detection sensor is designed and sustainability observer suggested using Lyapunov stability theory has been investigated.