Robotic Guide Dog

A quadruped robotic guide that helps blind and visually impaired (BVI) people navigate complex indoor & outdoor spaces—safely, smoothly, and with less cognitive load.

Assumptions

1. BVI Users are already familiar with basic assistive tools (white cane, smart cane).
2. The primary navigation contexts are indoor crowded spaces (cafeterias, malls, stations) and semi-outdoor areas (mixed terrain, curbs, lawns).
3. Trust and cognitive load are as important as speed and safety.

The Problem Discovered by Interviews

Existing tools (white cane, smart cane) help with obstacle detection but break down in unstructured environments. They:

1. Provide late or excessive feedback, creating high cognitive load.
2. Struggle with mixed terrains and overhanging obstacles.
3. Place full decision-making burden on the user.

The Solution: RDog

A quadruped robotic guide designed to proactively lead users, not just alert them. Key Features:

1. Autonomous wayfinding in both structured corridors and open layouts.
2. Proactive obstacle avoidance using LiDAR + onboard compute.
3. Kinesthetic guidance via a rigid, adjustable leash.
4. Preemptive voice cues for turns, terrain shifts, curbs, stairs.
5. Simple user control: pause, resume, adjust speed with joystick & leash-force inputs.

Market & User Fit

1. Users: BVI individuals seeking independence in daily navigation.
2. Market: Global assistive technology market (~$30B+ and growing).
3. Differentiator: Moves beyond “detection tools” to trusted, guided mobility.

Impact

Real-world studies with BVI participants show:

1. Navigation speed: 83.9s avg (RDog) vs. 125.9s (white cane) / 144.3s (smart cane). 2. Zero interventions required during trials.
3. Near-zero collisions across environments.
4. Lower workload: significantly reduced NASA-TLX scores.
5. High usability: SUS on par with white cane, exceeding smart cane.
6. User trust: Preemptive voice cues + slowdowns rated as critical for confidence.

Shortcomings & Trade-offs

1. Current system requires user to learn basic joystick inputs.
2. Battery and portability constraints limit all-day use.
3. Outdoor deployment in heavy rain/complex weather still under development.

Roadmap

1. Near term (6–12 months):
- Expand voice command library (“find a seat,” “return tray”).
- Improve all-weather resilience and portability.
2. Mid-term (12–24 months):
- Integrate crowd-aware pathfinding for busy pedestrian streets.
- Add safety behaviors modeled on guide-dog best practices.
3. Long term (24+ months):
- Scale to mass production with customizable options for BVI communities globally.
- Explore integrations with smart cities & transport hubs.