Trajectory tracking with high accuracy is a very challenging topic in direct drive robot control. This is due to the nonlinearities and input couplings present in the dynamics of the arm. Modelling difficulties like these have forced to use simplified or linearised models. Currently, one widely adopted approach in control engineering to overcome these modelling difficulties involves robust control, such as the sliding mode control. However there is a serious lack of effective or tractable in these control design methods since the designs mostly based on trial-and-error simulations. This paper presents the development of an integral sliding mode controller for the tracking problem of a 2 DOF direct drive robot arm based on genetic algorithm automated approach. A robust integral sliding mode controller is derived so that the actual trajectory tracks the desired trajectory as closely as possible.