Yupe — AI-Powered Accessible Journey Planner

Design Research & Proposal, Microsoft (Feb–March 2025)

A voice-assisted transit app concept, developed and presented to Microsoft's board as part of a research proposal into AI-powered experiences.

I led a team of five through the project — the app concept was developed collaboratively as a team, and I personally designed the brand (logo and name, "Yupe") and the full UI/UX across all six core screens in Figma. I then presented the final concept and working prototype directly to Microsoft's board, covering the vision, the mock-up, its potential impact, and future opportunities.

The Problem

London's transport network isn't consistently accessible — step-free access can be unreliable or hard to verify in the moment, and people with mobility needs often only discover a barrier once they've already arrived. Yupe explores how an AI-powered voice assistant could close that gap in real time, letting users report and check accessibility barriers hands-free, without breaking their journey.

  • Sign In


    Low-friction onboarding via standard SSO, minimising barriers to entry itself.

  • Navigation Quick-access actions

    (Route Planner, Report Barrier, Alerts, Statistics) sit directly over the live map, keeping core functions one tap away mid-journey.

  • Route Planner

    Route planner routes are presented with a transparent trade-off (faster vs. more accessible), putting the decision in the user's hands.

  • Social Feed


    A community layer turns individual accessibility reports into a shared, trustworthy live resource.

  • Voice Command

    Record Barrier The core AI interaction: users report an accessibility barrier hands-free via voice, critical for people who may already be navigating a physical obstacle.

  • Alerts


    The AI assistant surfaces real-time, community-reported issues — like step-free access problems at Waterloo — as a live "listening" interaction.

Next

Lee Kum Kee Advert