Simulation-based design and preliminary controller-level assessment of the real-time control logic of a proposed multimodal vibrotactile navigation belt

Abstract

Wearable assistive technologies for blind and visually impaired users require discreet feedback channels that do not rely primarily on continuous auditory output. This study presents a simulation-based proof of concept focused on the controller logic of a proposed multimodal vibrotactile navigation belt combining ultrasonic short-range obstacle detection, directional haptic feedback, and a discreet audio channel for high-priority events. The proposed architecture includes three HC-SR04 sensors, a light sensor, and a potentiometer for adjusting the distance threshold. The firmware, implemented in C on an STM32 microcontroller, uses separate threads for data acquisition, distance computation, event handling, and timing. The controller decision logic implemented in the simulation reference model was verified through deterministic unit testing, achieving 100% branch coverage. A Python simulator was then used to assess closed-loop controller–agent behavior in two-dimensional environments with obstacles and light variations. Across 30 base seeds and 30,000 episodes per condition, the mean collision rate was 0.978 ± 0.004 in the no-feedback baseline and 0.101 ± 0.008 in the controller-assisted condition, corresponding to a mean relative reduction of 89.70% ± 0.81%. These findings support controller-level technical feasibility but do not establish real-world usability, user effectiveness, or clinical benefit. Physical prototyping, calibration, and end-user evaluation remain necessary.