Sparse Swarm Research Articles

Exploring the Use of Terrestrial Robots for Atmospheric Electricity Measurement

Measuring the electric field is a central goal in electrostatics research, such as the study of electrostatics in atmospheric processes. It serves as a key indicator for various atmospheric phenomena, including the presence of lightning, dust or charged clouds. Traditionally, electric field measurements have been conducted from static platforms, with limited mobile measurements from airborne platforms such as balloons or, more recently, Unmanned Aerial Vehicles (UAVs). Here, we explore the potential of terrestrial robots to measure the electric field with some level of autonomy, such as during supervised navigation between user-defined waypoints. We mount a field mill on a four-wheeled rover and use a Global Navigation Satellite System (GNSS) to track the position of its measurements during an outdoor survey. The robot has a depth camera for 3D terrain mapping to contextualise local field measurements. We present plans for future research, including the use of semi-autonomous identification and exploration of electrostatic ‘hotspots’; and deployment of multiple robots (e.g., six) in a ‘sparse swarm’ configuration. We consider opportunities to employ such a robot system in environmental science or space research, for instance on Mars or the moon, where an understanding of electrostatic processes could be significant for future space missions.

Clustering Sparse Swarm Decomposition for Automated Recognition of Upper Limb Movements From Nonhomogeneous Cross-Channel EEG Signals

Clustering Sparse Swarm Decomposition for Automated Recognition of Upper Limb Movements From Nonhomogeneous Cross-Channel EEG Signals

Spaces between insects in laboratory swarms move like insects in natural swarms

Sparse swarms of flying insects show a high degree of spatial cohesion and are a form of collective animal behaviour; albeit one different from flocks and schools as they do not display ordered collective movements and under quiescent (laboratory) conditions long-range correlations are also absent. A better understanding of these outliers of collective behaviour may help to answer a long-standing open question in collective behaviour studies, namely: What is the signature that a group is “collective”? Even though dilute swarms of flying insects are mostly empty space no studies have reported on the dynamics of the spaces between swarming insects. Here I show that the spaces between insects (i.e., the centroids of empty tetrahedra formed by individuals and their 3 nearest neighbours) in laboratory swarms exhibit long-range (maximal) correlations and novel dynamic scaling in common with insects in natural swarms. Spaces within laboratory swarms therefore move like insects in natural swarms. I thereby unify seemingly disparate behaviours as long range correlations between individuals are absent in laboratory swarms but present in natural swarms. With the aid of stochastic trajectory models of non-interacting insects I show that long-range (maximal) correlations and the novel dynamic scaling arise generally and are not indicative of fine tuning. These results call for a re-evaluation of the importance of correlations and scaling in collective behaviours.

Neural-Swarm2: Planning and Control of Heterogeneous Multirotor Swarms Using Learned Interactions

We present <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Neural-Swarm2</i> , a learning-based method for motion planning and control that allows heterogeneous multirotors in a swarm to safely fly in close proximity. Such operation for drones is challenging due to complex aerodynamic interaction forces, such as downwash generated by nearby drones and ground effect. Conventional planning and control methods neglect capturing these interaction forces, resulting in sparse swarm configuration during flight. Our approach combines a physics-based nominal dynamics model with learned deep neural networks with strong Lipschitz properties. We make use of two techniques to accurately predict the aerodynamic interactions between heterogeneous multirotors: 1) Spectral normalization for stability and generalization guarantees of unseen data and 2) heterogeneous deep sets for supporting any number of heterogeneous neighbors in a permutation-invariant manner without reducing expressiveness. The learned residual dynamics benefit both the proposed interaction-aware multirobot motion planning and the nonlinear tracking control design because the learned interaction forces reduce the modelling errors. Experimental results demonstrate that <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Neural-Swarm2</i> is able to generalize to larger swarms beyond training cases and significantly outperforms a baseline nonlinear tracking controller with up to three times reduction in worst-case tracking errors.

Sparse Robot Swarms: Moving Swarms to Real-World Applications.

Robot swarms are groups of robots that each act autonomously based on only local perception and coordination with neighboring robots. While current swarm implementations can be large in size (e.g., 1,000 robots), they are typically constrained to working in highly controlled indoor environments. Moreover, a common property of swarms is the underlying assumption that the robots act in close proximity of each other (e.g., 10 body lengths apart), and typically employ uninterrupted, situated, close-range communication for coordination. Many real world applications, including environmental monitoring and precision agriculture, however, require scalable groups of robots to act jointly over large distances (e.g., 1,000 body lengths), rendering the use of dense swarms impractical. Using a dense swarm for such applications would be invasive to the environment and unrealistic in terms of mission deployment, maintenance and post-mission recovery. To address this problem, we propose the sparse swarm concept, and illustrate its use in the context of four application scenarios. For one scenario, which requires a group of rovers to traverse, and monitor, a forest environment, we identify the challenges involved at all levels in developing a sparse swarm—from the hardware platform to communication-constrained coordination algorithms—and discuss potential solutions. We outline open questions of theoretical and practical nature, which we hope will bring the concept of sparse swarms to fruition.

The impact of agent density on scalability in collective systems: noise-induced versus majority-based bistability

In this paper, we show that non-uniform distributions in swarms of agents have an impact on the scalability of collective decision-making. In particular, we highlight the relevance of noise-induced bistability in very sparse swarm systems and the failure of these systems to scale. Our work is based on three decision models. In the first model, each agent can change its decision after being recruited by a nearby agent. The second model captures the dynamics of dense swarms controlled by the majority rule (i.e., agents switch their opinion to comply with that of the majority of their neighbors). The third model combines the first two, with the aim of studying the role of non-uniform swarm density in the performance of collective decision-making. Based on the three models, we formulate a set of requirements for convergence and scalability in collective decision-making.

Balancing indoor thermal comfort and energy consumption of ACMV systems via sparse swarm algorithms in optimizations

This paper proposes a systematic modelling and optimizing of energy consumption and indoor thermal comfort for air-conditioning and mechanical ventilation (ACMV) systems. The models of extreme learning machines (ELM) and neural networks (NN) are established and evaluated. These well-trained models are then integrated with the computational intelligence techniques of sparse firefly algorithm (sFA) and sparse augmented firefly algorithm (sAFA). The sFA and sAFA aim to locate the global optimal operating points of the ACMV systems in real-time and predict energy saving rate (ESR) with a third order polynomial regression based on minimizing the mean squared errors (MSE) of the cost functions. This study also covers different indoor scenarios, such as general offices, lecture theatres and conference rooms. Given the well trained models, the maximum prediction of potential ESR can be −30% via the sparse AFA optimizations while maintaining indoor thermal comfort in the pre-defined comfort zone.

Observations of Mating Behaviour in the Honeybee

SUMMARYAn apparatus is described by which tethered virgin queens could be suspended for restricted flight at desired heights up to 11 m., enabling observations on mating. Drones were attracted in large numbers and seemed to show normal mating behaviour. Their flight near the queen is described in detail, and the methods by which they locate her. Drones, first attracted by the scent of queen sex attractants, approached from windward, flying typically in sparse swarms that assembled and hovered in conical formation below and behind the tethered queens. Drones were consistently observed to mount on top of the queen's abdomen, but successful mating followed rather rarely. Mating behaviour was very stereotyped; the drone clasped the queen's abdomen and everted the genitals into the sting chamber, which must be open. Paralysis of the drone accompanied the onset of genital eversion; this caused the drone to swing over backwards, attached by the genitals which everted more completely as the physical restraints inherent in the initial position were released. Finally, as the drone hung from the queen, the mating act was terminated by an audible snap, presumably caused by compressed air in the drone genitals, and apparently serving to separate the drone physically from the queen. Mating required only a few seconds, and in normal circumstances the entire act probably takes place in the air.Many free-flying drones were induced to ‘mate’ with queens in which an open sting chamber was simulated by the removal of segments 7–10, leaving an open body cavity. Lack of success in getting routine matings, using either the tethering technique or the release of free-flying queens, was probably due to failure on the part of the virgin queens to open their sting chambers.Drone behaviour must be considered carefully when conducting bioassays of sex attractants. Weakly attractive fractions are difficult to detect unless drones are attracted into the neighbourhood by a lure of crude queen lipid extract, or a tethered virgin queen.