Understanding the Power of Control in Autonomous Vehicles for People with Vision Impairment
Robin N. Brewer, Vaishnav Kameswaran · 2018 · Proceedings of the 20th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2018) · doi:10.1145/3234695.3236347
Summary
This University of Michigan study investigates how people with vision impairments perceive control in semi-autonomous versus fully autonomous vehicles, a distinction previous research had not explored. The researchers conducted design-based focus groups with 15 participants (average age 59) who were blind or had low vision, recruited through the National Federation of the Blind and local channels. The study used SAE International's classification system to distinguish semi-autonomous (level 3, where the driver may need to intervene) from fully autonomous (level 5, no driver intervention needed) vehicles. Each focus group included a design component where participants created physical prototypes of voice-based and tactile solutions to problems they identified, using craft materials like popsicle sticks, clay, rubber bands, and pipe cleaners. The participant demographics were diverse in terms of vision loss — ranging from born blind to those who lost vision later in life — and notably, 7 of 15 participants had prior driving experience, with several revealing they still drive to some degree despite being legally blind. This driving experience profoundly shaped their preferences, with former drivers preferring semi-autonomous vehicles that preserved some sense of control, while those who had never driven preferred fully autonomous systems. The study frames its analysis through Bandura's concept of agency, examining intentionality, forethought, self-reactiveness, and self-reflectiveness as design considerations for autonomous vehicle interfaces.
Key findings
The study reveals that control and independence are deeply intertwined but distinct concerns for people with vision impairments in the context of autonomous vehicles. Participants expressed strong emotional connections to driving as a symbol of independence, with some spending over $1,000 monthly on alternative transportation. While autonomous vehicles were seen as potentially transformative for independence, trust was a dominant concern — participants worried about vehicle malfunctions, over-trusting the technology, and not understanding its boundaries. A key finding was the identity challenge: participants struggled to envision themselves as operators of autonomous vehicles, suggesting that an identity shift is needed for blind people to see themselves as having agency over, not just being passengers in, these vehicles. In the design sessions, participants overwhelmingly drew on familiar assistive technology metaphors to imagine vehicle interfaces. Tactile solutions were modeled after white canes and refreshable Braille displays — one participant created a dashboard device mimicking a probing cane for pothole detection, while another built a vibration system indicating nearby vehicles. Voice solutions were modeled after screen readers and Siri/GPS interactions, with participants scripting natural conversational exchanges with the vehicle. Notably, when designing tactile solutions for situational awareness, no group incorporated voice or audio despite being told they could, suggesting participants see tactile feedback as most appropriate for real-time environmental awareness and self-reflection about surroundings.
Relevance
This paper makes a critical contribution to accessible transportation design by showing that control is not binary — it exists on a spectrum that must match individual needs, experiences, and comfort levels. For accessibility practitioners, the most actionable insight is the power of metaphor-based design: participants designed interfaces by extending concepts they already knew from assistive technologies (white canes, Braille displays, screen readers, GPS), suggesting that autonomous vehicle interfaces for blind users should build on these familiar interaction patterns rather than inventing entirely new paradigms. The study also raises important questions about agency and identity in automated systems that extend beyond vehicles — as AI systems increasingly make decisions on behalf of disabled users, the tension between independence (having control) and convenience (not needing control) will become a central design challenge. The finding that prior driving experience significantly shapes preferences underscores the importance of not treating "people with vision impairments" as a monolithic group in design research.
Tags: blindness · low vision · autonomous vehicles · accessible transportation · independence · agency · participatory design · voice interface · haptic technology
Standards referenced: SAE J3016