Project Overview
This graduation project explores how AI-powered virtual assistants can enhance independence and productivity for people who are blind or have low vision (PBLV) in educational and professional environments. Despite progress in assistive technologies, everyday tasks such as navigating dynamic indoor spaces or using touchscreen devices still create friction and dependence on others.
The project was conducted in collaboration with Envision Technologies BV, with stakeholders including People who are Blind or have Low Vision (PBLV), Envision’s product  team, broader institutional actors like universities and workplaces (the context) and fulfilling the academic requirements.
A mixed-method, iterative approach was used: desk research, user interviews (11 participants), shadowing, co-creation workshops, and testing of existing AI vision models. Iterative prototyping and Wizard-of-Oz methods were then applied to evaluate ideas with real users.
From research, 19 issues were identified. Two critical challenges emerged:
A. Navigating to non-fixed indoor locations (e.g., meeting rooms, classrooms)
B. Using inaccessible touchscreens (e.g., coffee machines, booking panels)
User testing and value vs. complexity mapping revealed that indoor navigation was more frequent and impactful. Insights showed that effective solutions must be hands-free, user-paced, and anchored in sensory cues (e.g., sound, floor texture), with clear, step-based instructions.
The final design, Ally SenseScape, is a voice-activated indoor navigation assistant integrated with Envision Glasses. It combines floor plans, reference images, compass data, and environmental anchors to provide step-based, clock-oriented guidance. Users control pacing through simple commands, while fallback strategies handle uncertainty (e.g., missed turns, poor connectivity).
User evaluation, including one real blind participant and two blindfolded design students, confirmed that Ally SenseScape increases confidence, supports independent movement, and holds strong potential for scalable implementation in offices and universities
Back to Top