Discovery and Exploration of Novel Swarm Behaviors Given Limited Robot Capabilities

  title={Discovery and Exploration of Novel Swarm Behaviors Given Limited Robot Capabilities},
  author={Daniel S. Brown and Ryan Turner and Oliver Hennigh and Steven Loscalzo},
Emergent collective behaviors have long interested researchers. [] Key Method Our approach uses novelty search to explore the space of possible behaviors in an objective-agnostic manner. Given this set of explored behaviors we use dimensionality reduction and clustering techniques to discover a finite set of behaviors that form a taxonomy over the behavior space. We apply our methodology to a single, binary-sensor capability model. Using our approach we are able to re-discover cyclic pursuit and aggregation…

Coevolving behavior and morphology of simple agents that model small-scale robots

This study introduces a new standpoint to the well-studied predator-prey pursuit problem (PPPP) using an implementation of very simple predator agents, using nano-robots designed to be morphologically simple.

Shepherding with robots that do not compute

The findings show that the shepherding problem does not fundamentally require arithmetic computation or memory to be solved, and the obtained controller solution is robust with respect to sensory noise, and copes well with changes in the number of sheep.

Coevolution of the Asymmetric Morphology and the Behaviour of Simple Predator Agents in Predator-Prey Pursuit Problem

This study introduces a new standpoint to the well-studied predator-prey pursuit problem using an implementation of very simple predator agents intended to model the small-scale (micro and nano) robots.

A Minimalistic Approach to Segregation in Robot Swarms

A decentralized algorithm to achieve segregation into an arbitrary number of groups with swarms of autonomous robots is presented and a thorough analysis of the parameter space that enables this behavior to emerge, along with an analysis that explains the emergent behavior.

Evolution, Robustness and Generality of a Team of Simple Agents with Asymmetric Morphology in Predator-Prey Pursuit Problem

To enhance the performance of the evolved system of predator agents, an asymmetric morphology featuring an angular offset of the sensor, relative to the longitudinal axis is proposed, and experimental results show that this change brings a considerable improvement of both the efficiency of evolution and the effectiveness of the evolve capturing behavior of agents.

Finding Consensus Without Computation

A novel control strategy for solving swarms of agents that is the first to be free of arithmetic computation, robust against sensory noise, scalable to large swarm sizes, and generalizes to the problems of choosing between more than two options or between unequal options.

Circle Formation with Computation-Free Robots Shows Emergent Behavioural Structure

This paper demonstrates how behavioural structure, such as a finite state machine, can emerge in minimal robots without computation nor memory capabilities, and provides theoretical proof for its emergence and self-healing properties.

Spatial Coverage Without Computation

The deterministic controller, obtained through off-line optimization, achieves around 71–76% coverage in a test scenario with no robot redundancy, which corresponds to a 26–39% reduction of the area that is not covered, when compared to an optimized random walk.

Task-Agnostic Evolution of Diverse Repertoires of Swarm Behaviours

It is shown that the repertoires contain a wide diversity of swarm behaviours, and for most of the tasks, the behaviours in the repertoire have a performance close to the performance achieved by task-specific evolution.



Limited bandwidth recognition of collective behaviors in bio-inspired swarms

A novel framework for classifying the collective behavior of a bio-inspired robot swarm using locally-based approximations of a swarm's global features and a classifier that leverages local agent-level features to accurately recognize these collective swarm behaviors while sampling from only a small number of agents is presented.

Classifying swarm behavior via compressive subspace learning

This paper demonstrates improvement in performance over prior works with respect to limited agent samples over a wide range of agent models and scenarios, and extends previous work on subspace estimation and identification from missing data to that of compressive measurements, whereCompressive measurements arise due to agent positions scattered throughout the domain.

Evolving and Controlling Perimeter, Rendezvous, and Foraging Behaviors in a Computation-Free Robot Swarm

Previous work on computation-free swarm behaviors is extended and it is shown that it is possible to evolve simple control policies to form a perimeter around a target, rendezvous to a specific location, and perform foraging.

Kilobot: A low cost scalable robot system for collective behaviors

Kilobot is a low-cost robot designed to make testing collective algorithms on hundreds or thousands of robots accessible to robotics researchers, and allows a single user to easily operate a large Kilobot collective.

Human-Swarm Interactions Based on Managing Attractors

It is claimed that managing attractors theoretically allows a human to abstract the details of individual agents and focus on managing the collective as a whole and uses quorum sensing to manage trade-offs between the scalability of interactions and mitigating the vulnerability of the swarm to agent failures.

Collective States, Multistability and Transitional Behavior in Schooling Fish

This study demonstrates that collective motion can be effectively mapped onto a set of order parameters describing the macroscopic group structure, revealing the existence of at least three dynamically-stable collective states; swarm, milling and polarized groups.

Evolution of swarm robotics systems with novelty search

This study shows that novelty search is a promising alternative for the evolution of controllers for robotic swarms with lower complexity than fitness-based evolution, and can find a broad diversity of solutions for the same task.

Evolving Aggregation Behaviors in Multi-Robot Systems with Binary Sensors

This paper investigates a non-traditional sensing trade-off in swarm robotics: one in which each robot has a relatively long sensing range, but processes a minimal amount of information. Aggregation

Behavioral diversity with multiple behavioral distances

Using several BSM in a single run increases the performance while avoiding the need to choose the most appropriate BSM and switching between BSMs leads to better results than taking the mean behavioral diversity, while requiring less computational power.

Clustering objects with robots that do not compute

The sensor/controller solution was implemented on a physical system of 5 e-puck robots, and favourable results have been obtained in systematic experiments with 20 objects, with 86.5% of the objects being gathered into one cluster after 10 minutes (average across 10 trials).