From animal collective behaviors to swarm robotic cooperation

National Science Review, Journal Year: 2023, Volume and Issue: 10(5)

Published: Feb. 16, 2023

The collective behaviors of animals, from schooling fish to packing wolves and flocking birds, display plenty fascinating phenomena that result simple interaction rules among individuals. emergent intelligent properties the animal behaviors, such as self-organization, robustness, adaptability expansibility, have inspired design autonomous unmanned swarm systems. This article reviews several typical natural introduces origin connotation intelligence, gives application case behaviors. On this basis, focuses on forefront progress bionic achievements aerial, ground marine robotics swarms, illustrating mapping relationship biological cooperative mechanisms cluster Finally, considering significance coexisting-cooperative-cognitive human-machine system, key technologies be solved are given reference directions for subsequent exploration.

Language: Английский

---

Biologically inspired herding of animal groups by robots

Methods in Ecology and Evolution, Journal Year: 2023, Volume and Issue: 14(2), P. 478 - 486

Published: Jan. 2, 2023

Abstract A single sheepdog can bring together and manoeuvre hundreds of sheep from one location to another. Engineers ecologists are fascinated by this herding because the potential it provides for ‘bio‐herding’: a biologically inspired animal groups robots. Although many algorithms have been proposed, most studied via simulation. There variety ecological problems where management wild is currently impossible, dangerous and/or costly humans manage directly, which may benefit bio‐herding solutions. Unmanned aerial vehicles (UAVs) now deliver significant benefits economy society. Here, we suggest use UAVs bio‐herding. Given their mobility speed, be used in wide range environments interact with at sea, over land air. We present roadmap achieving using pair UAVs. In our framework, UAV performs ‘surveillance’ groups, informing movement second that herds them. highlight promise flexibility paired approach while emphasising its practical ethical challenges. start describing types experiments data required understand individual collective responses Next, describe how develop appropriate algorithms. Finally, integration into software hardware architecture.

Language: Английский

---

3D-MuPPET: 3D Multi-Pigeon Pose Estimation and Tracking

International Journal of Computer Vision, Journal Year: 2024, Volume and Issue: 132(10), P. 4235 - 4252

Published: May 7, 2024

Abstract Markerless methods for animal posture tracking have been rapidly developing recently, but frameworks and benchmarks large groups in 3D are still lacking. To overcome this gap the literature, we present 3D-MuPPET, a framework to estimate track poses of up 10 pigeons at interactive speed using multiple camera views. We train pose estimator infer 2D keypoints bounding boxes pigeons, then triangulate 3D. For identity matching individuals all views, first dynamically match detections global identities frame, use tracker maintain IDs across views subsequent frames. achieve comparable accuracy state art terms median error Percentage Correct Keypoints. Additionally, benchmark inference with 9.45 fps 1.89 3D, perform quantitative evaluation, which yields encouraging results. Finally, showcase two novel applications 3D-MuPPET. First, model data single results estimation 5 pigeons. Second, show that 3D-MuPPET also works outdoors without additional annotations from natural environments. Both cases simplify domain shift new species environments, largely reducing annotation effort needed tracking. best our knowledge 2D/3D trajectory both indoor outdoor environments individuals. hope can open opportunities studying collective behaviour encourages further developments multi-animal

Language: Английский

---

Emergence of splits and collective turns in pigeon flocks under predation

Royal Society Open Science, Journal Year: 2022, Volume and Issue: 9(2)

Published: Feb. 1, 2022

Complex patterns of collective behaviour may emerge through self-organization, from local interactions among individuals in a group. To understand what behavioural rules underlie these patterns, computational models are often necessary. These have not yet been systematically studied for bird flocks under predation. Here, we study airborne homing pigeons attacked by robotic falcon, combining empirical data with species-specific model escape. By analysing GPS trajectories flocking individuals, identify two new escape: early splits and turns, occurring even at large distances the predator. examine their formation, extend an agent-based 'discrete' escape manoeuvre single initiator, namely sudden turn interrupting continuous coordinated motion Both turns this rule. Their relative frequency depends on angular velocity position initiator flock: sharp periphery lead to more than turns. We confirm association data. Our highlights importance discrete uncoordinated manoeuvres advocates systematic across species.

Language: Английский

---

A multi-scale review of the dynamics of collective behaviour: from rapid responses to ontogeny and evolution

Philosophical Transactions of the Royal Society B Biological Sciences, Journal Year: 2023, Volume and Issue: 378(1874)

Published: Feb. 20, 2023

Collective behaviours, such as flocking in birds or decision making by bee colonies, are some of the most intriguing behavioural phenomena animal kingdom. The study collective behaviour focuses on interactions between individuals within groups, which typically occur over close ranges and short timescales, how these drive larger scale properties group size, information transfer groups group-level making. To date, however, studies have focused snapshots, studying timescales up to minutes hours. However, being a biological trait, much longer important behaviour, particularly change their lifetime (the domain developmental biology) from one generation next evolutionary biology). Here, we give an overview across long, illustrating full understanding this animals requires more research attention its biology. Our review forms prologue special issue, addresses pushes forward development evolution encouraging new direction for research. This article is part discussion meeting issue ‘Collective through time’.

Language: Английский

---

Dynamics of collective motion across time and species

Philosophical Transactions of the Royal Society B Biological Sciences, Journal Year: 2023, Volume and Issue: 378(1874)

Published: Feb. 20, 2023

Most studies of collective animal behaviour rely on short-term observations, and comparisons across different species contexts are rare. We therefore have a limited understanding intra- interspecific variation in over time, which is crucial if we to understand the ecological evolutionary processes that shape behaviour. Here, study motion four species: shoals stickleback fish ( Gasterosteus aculeatus ), flocks homing pigeons Columba livia herd goats Capra aegagrus hircus ) troop chacma baboons Papio ursinus ). First, describe how local patterns (inter-neighbour distances positions), group (group shape, speed polarization) during differ each system. Based these, place data from within ‘swarm space’, affording generating predictions about contexts. encourage researchers add their own update space’ for future comparative work. Second, investigate intraspecific time provide guidance when observations made scales can result confident inferences regarding motion. This article part discussion meeting issue ‘Collective through time’.

Language: Английский

---

Snail-inspired robotic swarms: a hybrid connector drives collective adaptation in unstructured outdoor environments

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: April 29, 2024

Abstract Terrestrial self-reconfigurable robot swarms offer adaptable solutions for various tasks. However, most existing are limited to controlled indoor settings, and often compromise stability due their freeform connections. To address these issues, we present a snail robotic swarm system inspired by land snails, tailored unstructured environments. Our also employs two-mode connection mechanism, drawing from the adhesive capabilities of snails. The free mode, mirroring snail’s natural locomotion, leverages magnet-embedded tracks mobility, thereby enhancing adaptability efficiency. strong analogous response disturbance, vacuum sucker with directional polymer stalks robust adhesion. By assigning specific functions each our achieves balance between mobility secure Outdoor experiments demonstrate individual robots exceptional synergy within swarm. This research advances real-world applications terrestrial in

Language: Английский

---

Diffusion during collective turns in bird flocks under predation

Frontiers in Ecology and Evolution, Journal Year: 2023, Volume and Issue: 11

Published: July 11, 2023

Moving in groups offers animals protection against predation. When under attack, grouped individuals often turn collectively to evade a predator, which sometimes makes them rapidly change their relative positions the group. In bird flocks particular, quick reshuffling of flock members confuses challenging its targeting single individual. This confusion is considered be greater when internal structure group changes faster (i.e. ‘diffusion’ higher). Diffusion may increase individual birds with equal radii (same angular velocity) but not keep paths parallel (by adjusting speed). However, how diffusion depends on behaviour well known. grouping way they interact each other, referred as ‘alarmed coordination’ (e.g., reaction frequency or cohesion tendency), effect such collective turning unknown. Here, we aimed gain an understanding dynamics flocks. First, investigate relation between alarmed coordination and diffusion, developed agent-based model Second, test relates turns equal-radii parallel-paths, metric deviation from these two types. Third, studied predation empirically, by analysing GPS trajectories pigeons small pursued RobotFalcon. As measure used instability neighbours: rate closest neighbours member are changing. our simulations, showed that this increases size, frequency, topological range, tendency types counter-intuitive ways specifics. Empirically, less than starlings those parallel-paths. Overall, work provides framework for studying across species.

Language: Английский

---

Inferring social influence in animal groups across multiple timescales

Philosophical Transactions of the Royal Society B Biological Sciences, Journal Year: 2023, Volume and Issue: 378(1874)

Published: Feb. 20, 2023

Many animal behaviours exhibit complex temporal dynamics, suggesting there are multiple timescales at which they should be studied. However, researchers often focus on that occur over relatively restricted scales, typically ones more accessible to human observation. The situation becomes even when considering animals interacting, where behavioural coupling can introduce new of importance. Here, we present a technique study the time-varying nature social influence in mobile groups across scales. As case studies, analyse golden shiner fish and homing pigeons, move different media. By analysing pairwise interactions among individuals, show predictive power factors affecting depends timescale analysis. Over short relative position neighbour best predicts its distribution group members is linear, with small slope. At longer timescales, however, both kinematics found predict influence, nonlinearity increases, number individuals being disproportionately influential. Our results demonstrate interpretations arise from behaviour highlighting importance multiscale nature. This article part discussion meeting issue ‘Collective through time’.

Language: Английский

---

Perception of motion salience shapes the emergence of collective motions

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: June 5, 2024

Abstract Despite the profound implications of self-organization in animal groups for collective behaviors, understanding fundamental principles and applying them to swarm robotics remains incomplete. Here we propose a heuristic measure perception motion salience (MS) quantify relative changes neighbors from first-person view. Leveraging three large bird-flocking datasets, explore how this MS relates structure leader-follower (LF) relations, further perform an individual-level correlation analysis between past future change rate velocity consensus. We observe prevalence positive correlations real flocks, which demonstrates that individuals will accelerate convergence with who have higher MS. This empirical finding motivates us introduce concept adaptive MS-based (AMS) interaction model. Finally, implement AMS ~10 2 miniature robots. Swarm experiments show significant advantage enhancing smooth evacuations confined environments.

Language: Английский

---