Input/output (I/O) linearization is a model-based control method that receives much attention for many research studies in chemical industries. Despite many advances in control theory, the I/O linearization is not widespread for a real-time implementation when compared with other model-based techniques. It requires differentiation and inversion of the process model-which becomes cumbersome as the complexity of the model increases. Thus, this work presents the development and experiment of an embedded I/O feedback control device with a pilot process that emulates an exothermic reactor with nonminimum phase behavior. Algorithms of an approximate I/O linearization control system is truncated by using Taylor series expansion, and it is embedded into a field-programmable gate array control device. The effectiveness of the embedded approximate I/O controller is compared with a digital proportional-integral (PI) controller for a test with the pilot process with a hardwarein-the-loop concept. The results show that the embedded model-based controller provides high performance for the servo problem and disturbance rejection, compared with the PI controller. The proposed embedded model-based controller shows the application to apply the advance control method into the field-programmable gate array hardware as a stand-alone control device, and it can perform the computation of a complex algorithm.