Eye-gaze tracking is a common modality used by individuals with limited fine motor skills to control devices within their environment. This work proposes the use of an electrooculography-based human machine interface system that allows users to control smart devices without requiring a graphical user interface. Instead, the system leverages a person’s innate gaze interaction with devices in the environment and by recording the individual’s position and head orientation, determines the device the user wants to control. Seven different, simple eye gestures are then used for controlling the device to which the user is connected. This preliminary offline study assesses the feasibility of such a system by quantifying the device selection accuracy as well as the eye movement classification accuracy obtained across six subjects. Specifically, the users could select between seven different smart devices with an accuracy of \(97.57\% \pm 0.03\%\) while their eye gestures could be detected reliably with an accuracy of \(97.22\% \pm 0.03\%\) .