In this paper, a new method for designing a fractional-order proportional-integral-derivative (FO-PID) controller with an application to an induction motor drive is proposed. In the proposed method, the motor drive is modeled using an autoregressive with exogenous input (ARX) model whose parameters are experimentally identified using real I/O data. A genetic algorithm is then used to find the FO-PID parameters. To guarantee the robustness of the controller against load variations, minimax optimization is adopted. To validate the results, the proposed controller is applied to a real-life motor drive using a hardware-in-the-loop (HIL) simulator. The experimental results show that the proposed controller significantly improves the time response of the induction motor compared to a conventional PID controller. It also shows it is robust against motor load variations.