The aim of this paper is to identify and control a pneumatic servo drive in real-time environment. Obtaining the system's dynamic model accurately can be difficult once the pneumatic servo-system has been assembled since its highly nonlinear in nature, as a result, some difficulties in servo-pneumatic system modeling and control. In order to, overcome the complexity associated with the system nonlinearity, auto-regressive moving-average (ARMA) model is employed to identify the system's dynamic model in real-time environment. The advantages of this approach include high accuracy in the estimated model, low cost, and time reduction in controller design. The results acquired from the online experimental measured data are used to predict a discrete transfer function of the pneumatic servo system. The fourth-order model with one-step prediction shows the best performance compared with different order estimated model with varying sizes of step. Due to the highly nonlinearity of the system under study, two sophisticated controllers, PID-type fuzzy logic controller and Fractional order PID controller were chosen and designedusingthree optimization algorithms, namely particle swarm optimization (PSO), genetic algorithm (GA), grey wolf optimization (GWO).