Designing Inclusive Voice Control for Shared Autonomous Ride‑Hailing

At Volkswagen AG, I spent six months designing a voice-controlled drop-off system for shared autonomous vehicles. The users were people with upper-limb mobility impairments who couldn't use touch interfaces, but the design challenge was universal:How can voice control interfaces provide accessible, safe, dignified transportation for users with impairments, while improving experience for all users?

I built a load model to understand how cognitive and motor demands compound in real driving conditions, then designed a multi-turn voice dialogue system around that constraint. I ran a Wizard-of-Oz user test with 9 participants across 6 tasks. The system scored 79.4 on SUS and 29 on NASA-TLX workload. The most important finding was that time pressure mattered more than complexity.

The deliverables included full interaction specs, dialogue state diagrams, fallback logic documentation, and Figma component libraries for the voice UI states.