Controlling of real-time pressure process is more difficult because of its high sensitivity than its simulation environment. In some cases, the dead time associated with the process leads to non-linear and unpredicted behaviour. The conventional PI controller family is widely adopted for these processes because of their easy implementation, simple structure, and fewer tunable parameters. However, their performance is not effective during the changes in the system’s dynamic conditions like variation in set-point, high-frequency noise, and external disturbances due to the control signal deterioration. Therefore, in this paper, a novel controller is formed by hybridizing the robust fractional-order PI controller with the dead-time compensating smith predictor to handle the stochastic delay, noise, and peak overshoot reduction. The proposed controller maintains the same number of controller parameters possessed by the conventional PID controller; this gives easy tuning like the conventional ones. The step-response characteristics are compared with traditional PI, fractional-order PI, and predictive PI controllers to validate the proposed controller’s ability to handle the process noise, external disturbances, variable set-point, and system dead-time.