Mobile Application to Assist in the Evacuation of Blind or Visually Impaired People Using IoT and Azure AI Computer Vision

Gonzales J.; Laguerre F.; Wong L. · Conference of Open Innovation Association, FRUCT (2026) · DOI: 10.23919/fruct70069.2026.11506541

Resumen

The vulnerability of visually impaired individuals to natural disasters constitutes a critical public health issue, particularly in Peru, where this demographic faces high exposure to seismic risks and fires lacking adequate assistive tools. Consequently, this study proposes a technological solution based on a mobile application integrated with an Internet of Things bracelet and Azure Artificial Intelligence services to assist in safe evacuation and navigation. The methodology encompassed the development of hybrid architecture combining an ESP32 microcontroller with ultrasonic and thermal sensors for reactive edge detection, and cloud-based computer vision services for preventive environmental analysis. Validation was conducted through three experimental scenarios: route analysis, object identification, and fire identification. Results demonstrated the system's efficacy, successfully classifying routes based on obstacle density and triggering immediate multisensory alerts in response to proximity and temperature hazards. This confirms that the IoT-AI integration balances preventive precision with the necessary reactive response in emergencies.

Instituciones: Universidad Peruana de Ciencias Aplicadas