DroneNavigator: Using Leashed and Free-Floating Quadcopters to Navigate Visually Impaired Travelers
Mauro Avila Soto, Markus Funk, Matthias Hoppe, Robin Boldt, Katrin Wolf, Niels Henze · 2017 · Proceedings of the 19th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS '17) · doi:10.1145/3132525.3132556
Summary
This paper explores an unconventional approach to navigation assistance for visually impaired travelers: using small, lightweight quadcopters as mobile sound and haptic sources to guide users through indoor environments. While existing navigation systems deliver directional information through headphones, speakers, or vibrotactile actuators, these approaches use virtual or body-mounted feedback. DroneNavigator instead provides an actual moving sound source in 3D space — the quadcopter itself — that the user follows by listening to its rotor noise. The researchers developed two prototypes. The free-floating quadcopter flies approximately 1 meter in front of the user at shoulder height, providing auditory-only guidance through its natural rotor sound. It also signals elevation changes (stairs) by adjusting its altitude. The leashed quadcopter adds a physical tether connecting the drone to a custom handle held by the user, providing tactile directional feedback through the pull of the leash. The handle features a rotatable pane with a 3D-printed arrow that always points in the direction of travel, and the leash connection also powers an external 10,000mAh battery that extends flight time from 5-8 minutes to approximately 5 hours. Both prototypes use the small Eachine H8 Mini quadcopter (13.5x13.5cm, 20g). Navigation was controlled via Wizard of Oz in a controlled indoor study.
Key findings
In a within-subjects study with 14 visually impaired participants (13 completely blind, 1 with 5% remaining vision; ages 21-74, mean 51.69), all white cane users, three navigation conditions were compared across equally difficult 60-meter indoor routes containing staircases and doorways. The free-floating quadcopter was significantly faster (M=82.29s) than audio navigation (M=98.57s), with a large effect size (eta-squared=.254). Both quadcopter methods produced significantly fewer navigation errors than audio navigation (M=0.14 errors for both quadcopter types vs. M=0.86 for audio; large effect size eta-squared=.325). There was no significant difference in perceived cognitive workload (NASA-TLX) across conditions. In qualitative feedback, 8 of 14 participants preferred audio navigation overall due to familiarity, while 4 preferred free-floating and 2 preferred leashed quadcopter navigation. Participants valued that the free-floating quadcopter freed their hands for the white cane, and that the quadcopter provided continuous directional feedback unlike discrete audio instructions. However, participants raised concerns about the annoying rotor sound over extended use, social acceptability of drones as navigation aids, uncertainty about effectiveness in noisy environments, and the leashed version requiring both hands (one for handle, one for cane).
Relevance
DroneNavigator represents a creative exploration of how robotics might expand the palette of navigation aids available to blind travelers. The core insight — that a physical moving sound source in 3D space can convey directional information more naturally and continuously than verbal turn-by-turn instructions — has merit beyond the specific drone implementation. The finding that participants navigated faster and with fewer errors using quadcopters compared to state-of-the-art audio navigation is notable, even in this controlled setting. For accessibility researchers and practitioners, the paper raises important practical questions about the social dynamics of assistive technology in public spaces, drawing an explicit parallel to guide dogs: just as guide dogs required cultural acceptance, drone navigation aids would need similar normalization. The participants' concerns about noise, social stigma, and environmental limitations (wind, rain, noisy spaces) honestly highlight the gap between promising lab results and real-world deployment. The leashed prototype's clever solution of using the tether for both haptic feedback and battery power demonstrates resourceful engineering for accessibility applications.
Tags: navigation · blindness · drones · assistive technology · haptic technology · spatial audio · multimodal interaction · wayfinding