Examining the human-centred challenges of human-swarm interaction.

Abstract

We adopt a tripart approach in describing the human-centred challenges with human-swarm interaction. First, the results of large-N laboratory studies will be discussed which found evidence of trust biases (e.g. system-wide trust (SWT)), uncertainties regarding appropriate control strategies and timing, and inaccuracies of perceived swarm reliability. Second, the results of a small-N subject-matter expert study will be discussed, highlighting evidence for component-specific trust (CST) and will highlight concerns from expert operators regarding: understanding appropriate inputs for swarm control, transfer of control, general transparency for swarms including the need for understanding why errors might occur and concerns regarding the effects of command latency. Third, human-centred challenges from the use of swarms in the real world will be discussed. These challenges include lessons learned regarding human-swarm interfaces, the intense challenges of logistics when using swarms in the real world, and otherwise hidden challenges of using swarms for real-world scenarios such as the Defense Advanced Research Project Agency (DARPA) OFFensive Swarm-Enabled Tactics (OFFSET) programme and the use of swarms for monitoring operations for the Air Force Marathon. This article will close with a brief discussion of the opportunity space within human-swarm interaction if the research community is able to close the gap on these human-centred challenges.This article is part of the theme issue 'The road forward with swarm systems'.