ENHANCING NAVIGATION FOR THE VISUALLY IMPAIRED: A MAMDANI TYPE 1 FUZZY LOGIC APPROACH TO OBSTACLE DETECTION

The traditional white cane though commonly used by the visually impaired only has little usability in the complex and dynamic environment. Most previous research has focused on image detection algorithms for assistive devices, but they are less reliable in poor visibility and sensitive to environmental conditions. This study proposes an innovative solution through the combination of haptic feedback with a Fuzzy Logic Controller (FLC) for the better detection of obstacles and user notification. The proposed system incorporates MATLAB applications that enables the vibration intensity to be adjusted depending on the distance and direction of the obstacles and is more convenient for the user compared to the traditional methods. This work is new in the sense that it uses distance and direction as the two inputs in the fuzzy controller, which uses six fuzzy rules to control the vibration speed and direction of the actuator. This fine-tuned control allows the user to clearly identify the distance and location of the obstacles. The system's performance was validated with a root-mean-square-error (RMSE) of 0.568, which indicates a high level of accuracy in the model's predictions. This low RMSE is the result of small difference between the predicted and real obstacle distances and directions and proves that the system is reliable providing accurate feedback in real-life situations. Thus, this contribution is a valuable improvement in the field of assistive technologies for visually impaired, providing a more effective means for navigation by the visually impaired.