Smart Blind Stick with Machine Learning Based Obstacle Detection and Environmental Harzard Monitoring

Abstract

An intelligent assistive technology called the clever  Blind Stick with Machine Learning-Based Obstacle Detection and Environmental Hazard Monitoring is an intelligent assistive technology intended to increase the freedom, safety, and mobility of people with visual impairments. The system combines an Arduino-based embedded platform with a web-based dashboard, voice guidance, real-time sensor monitoring, machine learning, and emergency communication.By using three ultrasonic sensors to identify impediments in the front, left, and right directions, real-time speech advisories for safe navigation are made possible. Environmental risks including fire and damp surfaces are detected by a water level sensor and a flame sensor, which provide instant alerts. In the event of an emergency, a panic button can be used to notify designated contacts.The Arduino sends sensor data to a Python Flask backend, which uses machine learning models to interpret it and provide real-time data via an online dashboard. For efficient monitoring, the dashboard offers real-time sensor values, obstacle distances, prediction results, and alert history. To increase prediction accuracy and lower false alarms, the system uses a number of supervised machine learning algorithms with a majority voting mechanism, such as Random Forest, Decision Tree, K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Logistic Regression, and Linear Regression. By combining obstacle detection, hazard monitoring, voice guidance, emergency communication, and real-time visualization, the suggested smart blind stick provides a cost-effective, dependable, and intelligent mobility assistance solution, improving the safety and quality of life of visually impaired people