OpenGuide: A Scalable Human-Like Guidance System for Visually Impaired Travelers

Event: PAVE Inclusive Design Challenge · Duration: 14 min · ▶ Watch on YouTube

Abstract

The Boston University team presents OpenGuide, a scalable human-like guidance system designed to enhance mobility for visually impaired travelers. The project addresses the critical need for personalized and context-dependent navigation solutions, recognizing that individuals react uniquely to guidance. Through collaboration with industry partners and organizations for the blind, OpenGuide aims to provide a seamless and safe experience, empowering users by incorporating their feedback into the system’s continuous learning and adaptation.

Speakers

  • Tabitha Colter — PAVE
  • Eshed Ohn-Bar — Boston University
  • Matthew Boyd — Boston University
  • Andres Armijos — Boston University
  • Minglan Zheng — Boston University
  • Jim Zhang — Boston University

Talks (12)

  • 00:00:29 — Tabitha Colter: Introduction to the Boston University Team
    • Tabitha Colter introduces the Boston University team and their involvement in the Inclusive Design Challenge.
  • 00:00:37 — Eshed Ohn-Bar: Eshed Ohn-Bar’s Background and Research Focus
    • Eshed Ohn-Bar discusses his decade-long work in autonomous driving and assistive technologies, particularly for people with disabilities.
  • 00:01:46 — Matthew Boyd: Matthew Boyd’s Motivation for Robotics and Autonomous Systems
    • Matthew Boyd shares his transition from business IT to robotics and autonomous systems, driven by a desire to help people with mobility challenges.
  • 00:02:44 — Andres Armijos: Andres Armijos’s Journey to Human-Centered Robotics
    • Andres Armijos explains his background in aerospace engineering and localization systems, and his shift towards human-centered assistive technologies.
  • 00:03:55 — Minglan Zheng: Minglan Zheng’s Work on Pedestrian Simulation
    • Minglan Zheng describes her undergraduate research on simulating pedestrians with disabilities and their interaction with vehicles.
  • 00:05:11 — Matthew Boyd: Overview of the OpenGuide Project
    • Matthew Boyd introduces OpenGuide as a scalable assistive system providing seamless navigation for people with disabilities across various scenarios.
  • 00:06:39 — Eshed Ohn-Bar: Origin and Evolution of OpenGuide: Personalizing Guidance
    • Eshed Ohn-Bar discusses the unexpected finding that users react uniquely to the same guidance, leading to the core idea of personalizing the OpenGuide system.
  • 00:08:38 — Andres Armijos: Partnerships for OpenGuide Development
    • Andres Armijos highlights partnerships with Motional and The Carroll Center for the Blind, bringing industry perspective and invaluable resources to the project.
  • 00:09:40 — Minglan Zheng: Community Engagement and User Input
    • Minglan Zheng explains the team’s collaboration with The Carroll Center for the Blind for use case design, user recruitment, and prototype testing.
  • 00:10:22 — Jim Zhang: Challenges and Lessons Learned: Safety and User Diversity
    • Jim Zhang discusses the fundamental challenges of inclusive navigation, emphasizing safety and the need to understand and adapt to diverse user needs.
  • 00:11:26 — Matthew Boyd: Ultimate Aspiration for OpenGuide: Empowering Users
    • Matthew Boyd articulates the vision for OpenGuide to act as an expert aid, providing personalized awareness and empowering users to be active participants in the system’s development.
  • 00:12:41 — Minglan Zheng: Broader Vision: Inclusive Mobility Systems
    • Minglan Zheng emphasizes the project’s role in advancing inclusive mobility by making technology accessible to all, especially for pedestrians with disabilities at critical junctions.

Key Takeaways

  • OpenGuide aims to provide a universally accessible, personalized, and context-dependent navigation experience for people with diverse disabilities.
  • A key insight is that each user reacts uniquely to guidance, necessitating a system that learns and adapts in real-time to individual needs and preferences.
  • Safety is a paramount concern, especially as assistive devices transition from controlled environments to the real world, requiring robust solutions to prevent collisions and ensure user well-being.
  • The project emphasizes active user participation, viewing visually impaired individuals as engineers and designers who provide invaluable real-time feedback to improve the system’s scope, adaptability, and effectiveness.
  • Collaboration with industry leaders like Motional and community organizations like The Carroll Center for the Blind is crucial for integrating diverse perspectives and resources into the development of inclusive mobility solutions.

Methods / Models / Datasets Mentioned

  • OpenGuide
  • Uber
  • Lyft
  • Assistive Indoor Navigational Assistance Technologies
  • Localization Systems
  • Product-centered approach
  • Human-centered approach
  • Simulation of pedestrians with disabilities
  • Prototype testing

Topics

Autonomous Vehicles · Assistive Technology · Visual Impairment · Navigation Systems · Personalized Guidance · User-Centered Design · Accessibility · Human-Robot Interaction · Safety-Critical Applications · Inclusive Mobility

Notes

Open for commentary — connections to other work, critiques, follow-up reading.