Timing and Integration of Directional Cues

Abstract Robots that share environments with people should communicate their directional intentions so that people can anticipate their future directions. In the current paper, the aim is to find whether using multiple cues at the same time improves the predictability of the robot. In the first experiment, participants were shown videos of an animated robot communicating its future direction with a LED cue, a rotating screen with or without projected eyes and a combination of these cues. Results indicated that any cue diminished reaction times and increased the predictability of the robot and that a combination of cues scored even better. In the final experiment, we validated this finding using an autonomous robot in a lab setting. We compared a LED cue and the rotation of a screen and the combination of cues and also added the timing of this cue. Results indicated that the LED cue scored better on perceived predictability than the rotation of the screen and that multiple cues increased the predictability. Perceived comfort was influenced by robot speed, and predictability increased if the robot starts signalling earlier. The results show that robots can effectively combine cues and communicate earlier, and the current paper improves the predictability of robots in the field.