Adaptivity: a path towards general swarm intelligence?

The field of multi-robot systems (MRS) has recently been gaining increasing popularity among various research groups, practitioners, and a wide range of industries. Compared to single-robot systems, multi-robot systems are able to perform tasks more efficiently or accomplish objectives that are simply not feasible with a single unit. This makes such multi-robot systems ideal candidates for carrying out distributed tasks in large environments-e.g., performing object retrieval, mapping, or surveillance. However, the traditional approach to multi-robot systems using global planning and centralized operation is, in general, ill-suited for fulfilling tasks in unstructured and dynamic environments. Swarming multi-robot systems have been proposed to deal with such steep challenges, primarily owing to its adaptivity. These qualities are expressed by the system's ability to learn or change its behavior in response to new and/or evolving operating conditions. Given its importance, in this perspective, we focus on the critical importance of adaptivity for effective multi-robot system swarming and use it as the basis for defining, and potentially quantifying, swarm intelligence. In addition, we highlight the importance of establishing a suite of benchmark tests to measure a swarm's level of adaptivity. We believe that a focus on achieving increased levels of swarm intelligence through the focus on adaptivity will further be able to elevate the field of swarm robotics.

Adapting the Exploration-Exploitation Balance in Heterogeneous Swarms: Tracking Evasive Targets.

There has been growing interest in the use of multi-robot systems in various tasks and scenarios. The main attractiveness of such systems is their flexibility, robustness, and scalability. An often overlooked yet promising feature is system modularity, which offers the possibility of harnessing agent specialization, while also enabling system-level upgrades. However, altering the agents' capacities can change the exploration-exploitation balance required to maximize the system's performance. Here, we study the effect of a swarm's heterogeneity on its exploration-exploitation balance while tracking multiple fast-moving evasive targets under the cooperative multi-robot observation of multiple moving targets framework. To this end, we use a decentralized search and tracking strategy with adjustable levels of exploration and exploitation. By indirectly tuning the balance, we first confirm the presence of an optimal balance between these two key competing actions. Next, by substituting slower moving agents with faster ones, we show that the system exhibits a performance improvement without any modifications to the original strategy. In addition, owing to the additional amount of exploitation carried out by the faster agents, we demonstrate that a heterogeneous system's performance can be further improved by reducing an agent's level of connectivity, to favor the conduct of exploratory actions. Furthermore, in studying the influence of the density of swarming agents, we show that the addition of faster agents can counterbalance a reduction in the overall number of agents while maintaining the level of tracking performance. Finally, we explore the challenges of using differentiated strategies to take advantage of the heterogeneous nature of the swarm.

Beyond Bio-Inspired Robotics: How Multi-Robot Systems Can Support Research on Collective Animal Behavior.

In the study of collective animal behavior, researchers usually rely on gathering empirical data from animals in the wild. While the data gathered can be highly accurate, researchers have limited control over both the test environment and the agents under study. Further aggravating the data gathering problem is the fact that empirical studies of animal groups typically involve a large number of conspecifics. In these groups, collective dynamics may occur over long periods of time interspersed with excessively rapid events such as collective evasive maneuvers following a predator's attack. All these factors stress the steep challenges faced by biologists seeking to uncover the fundamental mechanisms and functions of social organization in a given taxon. Here, we argue that beyond commonly used simulations, experiments with multi-robot systems offer a powerful toolkit to deepen our understanding of various forms of swarming and other social animal organizations. Indeed, the advances in multi-robot systems and swarm robotics over the past decade pave the way for the development of a new hybrid form of scientific investigation of social organization in biology. We believe that by fostering such interdisciplinary research, a feedback loop can be created where agent behaviors designed and tested in robotico can assist in identifying hypotheses worth being validated through the observation of animal collectives in nature. In turn, these observations can be used as a novel source of inspiration for even more innovative behaviors in engineered systems, thereby perpetuating the feedback loop.

Balancing Collective Exploration and Exploitation in Multi-Agent and Multi-Robot Systems: A Review.

Multi-agent systems and multi-robot systems have been recognized as unique solutions to complex dynamic tasks distributed in space. Their effectiveness in accomplishing these tasks rests upon the design of cooperative control strategies, which is acknowledged to be challenging and nontrivial. In particular, the effectiveness of these strategies has been shown to be related to the so-called exploration-exploitation dilemma: i.e., the existence of a distinct balance between exploitative actions and exploratory ones while the system is operating. Recent results point to the need for a dynamic exploration-exploitation balance to unlock high levels of flexibility, adaptivity, and swarm intelligence. This important point is especially apparent when dealing with fast-changing environments. Problems involving dynamic environments have been dealt with by different scientific communities using theory, simulations, as well as large-scale experiments. Such results spread across a range of disciplines can hinder one's ability to understand and manage the intricacies of the exploration-exploitation challenge. In this review, we summarize and categorize the methods used to control the level of exploration and exploitation carried out by an multi-agent systems. Lastly, we discuss the critical need for suitable metrics and benchmark problems to quantitatively assess and compare the levels of exploration and exploitation, as well as the overall performance of a system with a given cooperative control algorithm.