FSE 100: AI Vision Project

As part of an Introduction to Engineering course, collaborated on the design and construction of a wearable assistance device intended to improve safety for visually impaired users. The goal was to address hazards above waist level—an area not detectable by traditional white canes and a common source of injury.

The system combined computer vision and proximity sensing to detect obstacles and provide real-time feedback to the user. A camera captured images of the surrounding environment, which were analyzed via an AI vision API to identify nearby obstacles. When an obstacle was detected, a vibration motor provided tactile feedback to warn the user.

In parallel, an ultrasonic sensor continuously measured distance to nearby objects. If an object entered a defined safety threshold, an audible buzzer was triggered to provide an additional alert. This multi-sensor approach improved reliability by combining AI-based recognition with direct distance measurement.

The final prototype demonstrated how layered sensing and feedback mechanisms can enhance user awareness and reduce the risk of head-level collisions.