Pneumatic actuators are the most commonly used final control elements in industrial processes and abnormal behavior of which impacts the quality of the end product. Hence detection and identification of pneumatic actuator faults are essential. In most of the works, the time of detection of faults and the severity levels of the faults are not reported, and also the inflow to the control valve is not considered as an influential parameter. In this work, an attempt is made to detect and diagnose two pneumatic actuator faults, stem displacement fault and insufficient supply pressure fault using a decision fusion process. The two parameters such as valve stem displacement and the valve output flow rate are considered for the process of fusion. The decision fusion is performed using a fuzzy-based inference system to detect and identify the severity of fault into low, medium, and high faults. The experiment is conducted on a flow process and the proposed method can detect the faults online. The proposed technique can identify the insufficient supply pressure fault within 1-2s of its occurrence with a 100% detection ratio whereas stem displacement fault is detected within 3-6s of its occurrence with a detection ratio of 93.75%.