Comparison of tactile signals for collision avoidance on unmanned aerial vehicles

Abstract

Our recent work focused on the development of intuitive user interfaces for the control of unmanned aerial vehicles, such as quadcopters. Next to intuitive gesture control, a key challenge with remotely operated quadcopters is the display of information about the aircraft surroundings. To this end, we examined the use of rendering tactile stimuli to warn about nearby obstacles. Directional information and distance is encoded via vibrotactile signals from rotating mass motors. Three different methods of delivering the tactile feedback were tested in a user study. Results show that even though participants guided the quadcopter through a maze by tactile stimuli alone, they were, on average, able to avoid full crashes. Further, we found that using sequential signals to indicate obstacles lead to significantly increased numbers of wall contacts.