A Mechatronic Eye Movement Emulator for Benchmarking AR/VR Eye-Tracking Sensors

Eye-tracking sensors for augmented and virtual reality (AR/VR) devices are rapidly evolving. To support standardized performance evaluation, we introduce an opensource eye movement emulator comprising: (a) an anthropomorphic head model, (b) bioinspired artificial eyes with anatomically accurate pupil, iris, and cornea, and (c) a gaze control system that automatically follows external visual targets. Experimental validation using commercial eye-tracking glasses with MEMS- and camera-based sensors demonstrated that the emulator is accurate and robust, reporting gaze accuracy comparable to prior tests with human subjects. This emulator provides a reproducible and automated benchmark tool for cross-platform eye-tracking sensor development.