Preferred Distance in Human-Drone Interaction

Abstract

In two augmented-reality experiments, we transferred the paradigm of interpersonal distance regulation to human-drone interaction. In the first experiment, we used a simple spherical drone model and explored how both hovering height and approach angle affect the preferred distance. Drone height above the ground had a strong effect. The preferred distance to the drone was larger than that typically found toward human actors, in particular, when the drone trajectory was very high. In the second experiment, we sought to gain a deeper understanding of the factors that may influence this effect. In addition to the simple spherical drone model used in the first experiment, we also varied its appearance and attachment to the ground. Surprisingly, anthropomorphic features increased preferred distances. We, therefore, discuss the extent to which social aspects and subjectively perceived danger influence the preferred distance for interaction with drones, which thus need to be considered in the design of human-drone interaction.

Keywords: drone appearance; human–drone interaction; interpersonal distance; personal space; proxemics.

Figure 1

Depiction of the projected drone in the room as seen through the HoloLens 2.

Figure 2

Experimental setup in Experiment 1. The subject stood at the marked point in the room, facing in the designated direction. The ceiling height was 3 m. The drone could be displaced along the respective marked trajectories and kept the same hovering height throughout a trial. The room was a common room/meeting area furnished with a small tea kitchen, tables, and chairs.

Figure 3

Procedure of the laboratory experiment.

Figure 4

Mean comfortable distance as a function of drone height. The grey dashed line represents the subjects’ eye level. Symbols differing in shape and color represent the approach angles. Error bars indicate ± 1 SE of the mean. The left panel (a) shows the averaged results of the three approach angles depicted in the right panel (b).

Figure 5

The projected drone models and the human male model in the room as seen through the HTC XR Elite: (a) Simple spherical form; (b) Variant with anthropomorphic features; (c) Variant with anthropomorphic features plus ground anchoring; (d) Human male model. All models are positioned with their anchor point at a height of 110% of eye level and at a distance of 2 m.

Figure 6

Experimental setup in Experiment 2. The subject stood at the marked location in the room, facing in the indicated direction. The room was a meeting room furnished with a table, chairs and writing surfaces.

Figure 7

Mean comfortable distance as a function of drone height. The grey dashed line represents the subjects’ eye level. Symbols differing in shape and color represent the different models. Error bars indicate ± 1 SE of the mean. The left panel (a) shows the averaged results of the three variants of the drone model depicted in the right panel (b).