More than the sum of its parts: assessing the coherence and expressivity of a robotic swarm Florent Levillain, David St-Onge, Elisabetta Zibetti and Giovanni Beltrame Abstract— The robotics community is considering the use of large groups of robots, also known as artificial swarms for applications in unknown and dynamic environments. In this context, swarms of robot will need to interact with users to accomplish their mission. Unfortunately, little is known about the users’ perception of group behavior and dynamics, as well as what is the best interaction modality for swarms. In this paper, we focus on the movement of the swarm as a group to convey information to a user: we believe that the interpretation of artificial states based solely on the motion can lead to promising natural interaction modalities. We define the expressivity of a movement as a metric to understand how natural, readable, or easily understandable such movement may appear. We then correlate expressivity with the control parameters for the distributed behaviour of the swarm. A user study confirms the relationship between inter-robot distance, temporal and spatial synchronicity, and the perceived expressivity of the robotic system. I. INTRODUCTION As robots make their way into our world, the number of application domains where they are likely to interact and cooperate with humans multiplies. Each of these domains constitutes an opportunity to develop a more natural and in- tuitive relationship, working on the robots capacity to detect social attitudes and adopt expressive stances. While social robotics has mostly focused on humanoid and zoomorphic robots, new forms of robots are entering the scene. Robot swarms are one of them, composed of large numbers of robots that can evolve in formation and adapt easily to multiple environments. The robustness of swarm systems comes mostly from their distributed and scalable control. For interaction with humans, what makes swarms special is that they have no defined physicality: they can adopt emerging configurations depending on environmental constraints, inter- nal policies and commands issued by a user [1]. This absence of predictable structure, and the necessity for an observer to consider multiple individuals, make it necessary to develop new methods of evaluation to qualify the interaction with these robots. Research on the affective reactions to robot swarms has only started [2]. So far we possess scanty information about how a swarm’s motion impacts a user’s emotional response [3]. Specifically, we do not know how the state Dr. St-Onge and Dr. Beltrame are with the Department of Computer and Software Engineering, ´ Ecole Polytechnique de Montr´ eal, Qu´ ebec Canada e-mail: (david.st-onge@polymtl.ca). Dr. Levillain is with Ensadlab-Reflective Interaction. ´ Ecole Nationale Sup´ erieure des Arts D´ ecoratifs, 75240 Paris Cedex 05, France. email: (florent.levillain@ensad.fr). Dr. Zibetti is with the CHART-LUTIN Laboratory, Universit´ e Paris 8, 93526 Saint Denis Cedex 02, France. attributed to a swarm (e.g. is it considered as a single entity, an aggregate of autonomous robots, an ephemeral forma- tion?) affects its perceived psychological traits (nervous, shy, aggressive, etc.), as well as the expressivity that may be attributed to its behaviour. Are the reactions to a robot swarm similar to those we can feel when observing a school of fish or a flock of birds? Is a robot swarm able to impress the sense of a collective movement organized towards a goal? To what extent an affective relationship can be established with an ensemble of robots? This paper preliminarily addresses the fundamental ques- tions on the cohesion and on the expressivity of a swarm, and how they are dependent on a defined set of parameters. In particular we examine how cohesion and expressivity allow humans to understand the swarm motion dynamics, and perceive it as a single behavioral entity, as opposed to a collection of moving objects. These questions are addressed with a user study on a small swarm of table-top robots. In the following, we describe related literature, define expressivity, detail our distributed control mechanism, and how it is related to expressivity measurements. II. RELATED WORK This paper relates to two bodies of knowledge that are still somewhat new in robotics: human-swarm interaction and the use of non-verbal communication from robots. Both have some key contributions on which we base this work and the assumptions used in our study. A. Human-Swarm Interaction Human-Swarm Interaction (HSI) differs from common Human-Robot Interaction (HRI) for the large numbers of units involved and because it heavily relies on inter-robot communication. The handful of HSI studies currently pub- lished focus on specific interface media, and very few actually study user reaction/perception. In addition, most are conducted in simulation, suffering from a reality-gap [3]. Podevijn et al. [4] successfully showed by experiments that the number of robots does not influence the cognitive load required from a user if the control is performed on the swarm as a whole. To convey information about swarm states, a flexible strat- egy is to use iconic representation that users can recognize without having to recall them, such as the top LEDs on each of the robots [5], or make the robots emit machinic sounds [6]. Note that the latter uses sounds to help the user be aware of a malfunction in the swarm, not to share high- level state information.