ThermalCane: Exploring Thermotactile Directional Cues on Cane-Grip for Non-Visual Navigation
Arshad Nasser, Kai-Ning Keng, Kening Zhu · 2020 · Proceedings of the 22nd International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2020) · doi:10.1145/3373625.3417004
Summary
This paper presents ThermalCane, a white-cane grip augmented with flexible Peltier thermoelectric modules that provide thermal directional cues for blind and visually impaired (BVI) navigation. While existing haptic navigation aids primarily use vibrotactile feedback (vibration motors), the authors argue that vibration has significant drawbacks for cane users: the natural turbulence from tapping the cane on the ground while walking can mask or interfere with vibrotactile signals, and vibration can be difficult to localize precisely when multiple motors are used. Thermal feedback offers an alternative sensory channel that is unaffected by walking-induced vibrations. ThermalCane uses flexible Peltier modules (23x79x2mm each) wrapped around a 3D-printed cylindrical grip that attaches to existing white canes. The system was tested in three configurations: 3, 4, and 5 modules arranged in equal spacing around the grip. Each module can produce hot (+3 degrees C/s) or cold (-3 degrees C/s) stimuli for 1.5 seconds, and the spatial position of the activated module on the grip maps to navigation directions. The system is driven by an Arduino Uno with H-bridge motor drivers and a 9V battery, connected to a GPS-equipped smartphone for real-world navigation. The directional mapping uses cold stimuli for the primary directions: top module = Go Forward, left = Turn Left, right = Turn Right, bottom = U-Turn, with hot feedback on top reserved for Stop.
Key findings
Two studies were conducted. Study 1 with 12 BVI participants (ages 25-35, 6 congenitally blind) evaluated thermal perception across the three grip configurations. The four-module configuration was selected as optimal, balancing accuracy (84.7% overall) with expressiveness — the three-module grip was slightly more accurate but offered fewer directional options, while the five-module grip had significantly lower accuracy and higher cognitive load. Cold stimuli were perceived significantly more accurately than hot stimuli (p<0.005) with shorter response times, and 11 of 12 participants preferred the four-module configuration. Study 2 compared ThermalCane against vibrotactile feedback during outdoor walking with 6 BVI participants (4 congenitally blind, 2 with 25% vision). Thermal feedback yielded significantly higher directional accuracy than vibrotactile feedback (Z=2.21, p<0.05), with perfect 100% accuracy for Go, Stop, and U-Turn thermal cues versus much lower vibrotactile scores (55% Go, 85% Stop, 73.33% U-Turn). Response times were comparable (thermal 2.96s vs vibrotactile 2.87s). NASA-TLX workload scores showed slightly lower task load for thermal cues. All 6 BVI participants expressed preference for thermal over vibrotactile directional cues, with one noting: "I don't have to put much of effort in understanding the thermal sensation as it is very intuitive. I can focus on my surrounding noises." Participants suggested combining thermal and vibrotactile cues in a single device for both navigation and obstacle detection.
Relevance
ThermalCane introduces thermal feedback as a practical new modality for assistive navigation, addressing a real limitation of the dominant vibrotactile approach: signal masking from the physical act of using a white cane while walking. For accessibility researchers and developers, this opens a design space where thermal cues handle directional navigation while vibration handles obstacle notification — a multimodal approach that several participants explicitly requested. The finding that cold stimuli outperform hot stimuli in both accuracy and comfort has direct design implications: cold-based thermal feedback should be the default for safety-critical navigation cues. Practical limitations include the impact of environmental temperature on perception (outdoor conditions varied from 23-32 degrees C), the challenge of using thermal cues while wearing gloves, potential issues with prolonged grip contact causing sweating, and individual variation in thermal sensitivity. The small sample sizes (N=12, N=6) and the controlled outdoor walking environment (open field) mean real-world urban navigation performance remains to be validated.
Tags: visual accessibility · navigation · haptic technology · white cane · assistive technology · blindness and low vision · wayfinding · tactile accessibility · wearable technology