Shoe-Based Interface Promoting Instinctive Avoidance Behavior in Poor Visibility Conditions Utilizing Averse Behavior

In disasters such as the Great East Japan Earthquake and water accidents, many residents do not take evacuation action even after warnings are given, or they are careless or negligent about the danger. Additionally, poor visibility is likely to lead to serious injuries and accidents such as falls and tumbles. In such a disaster or dangerous situation, warnings are a very effective way to warn people of danger but the final decision to evacuate or avoid is individual. Previous studies suggest a prevailing optimistic bias towards disasters among individuals, underscoring the importance of emotional factors, which often override scientific information, in promoting evacuation behavior. Traditional warnings typically leverage visual or auditory signals, which may not be effective under conditions of poor visibility. Vibrotactile warnings, on the other hand, do not rely on these senses and can effectively alert individuals to imminent danger. However, those remains unclear whether vibrotactile warnings can counteract the optimistic bias towards threats and foster emotional factors that stimulate evasion. In this study, we propose a shoe-based interface using vibrotactile sensation to promote instinctive danger avoidance and evacuation behavior in disasters and poor visibility. By harnessing the natural 'averse behavior' of living organisms, our interface incites instinctive danger avoidance via unpleasant vibrations and tactile sensations. We conducted an experiment to identify the location of a vibration motor and an experiment to identify the vibration pattern that causes unpleasant emotions. The results showed that a toe was appropriate for the former, and a vibration pattern with a constant vibration time and a decreasing stop time was appropriate for the latter. Then, to verify the effectiveness of the shoe-based interface produced based on the results as a warning, we conducted a comparison experiment with existing sound warnings. We evaluated subjects' avoidance success and emotional responses utilizing questionnaires and biological data. Two experiments were conducted: one without prior explanation and one with prior explanation. Results showed a slightly lower success rate in avoiding danger compared to the sound warning, but both warnings without prior explanations induced over 80％ of unpleasant and negative emotions. The percentages were 66.7% and 57.1% for the vibration and sound warnings, respectively, with prior explanation. In a post-experiment question, over 90％ of subjects expressed not wanting to proceed with either warning. Therefore, this interface is slightly inferior to existing sound warnings in terms of stopping walking, but it has similar functionality in inducing unpleasant emotions, not wanting to go on, and having difficulty becoming accustomed to the stimulus. Moreover, our interface has the potential to promote instinctive avoidance and evacuation behaviors by instilling a sense of crisis. While the effectiveness of this approach was demonstrated in VR scenarios, its real-world effectiveness is yet to be established. Additionally, we did not consider the strength of the vibration while investigating the mounting positions and patterns. Future research will aim to investigate the ideal strength and number of vibration motors to optimize the system's performance and conduct real-world testing to evaluate its efficacy.