Ergodic Specifications for Flexible Swarm Control: From User Commands to Persistent Adaptation

Prabhakar, Ahalya; Abraham, Ian; Taylor, Annalisa; Schlafly, Millicent; Popovic, Katarina; Diniz, Giovani; Teich, Brendan; Simidchieva, Borislava I.; Clark, Shane; Murphey, Todd D.

doi:10.15607/rss.2020.xvi.067

Cited by 20 publications

(15 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To this end, we use the ergodic metric from Section 4.3 as an objective to synthesize maximally ergodic trajectories for general nonlinear systems using tools from model-predictive control [204]. However, we note that any trajectory optimization tools or direct optimization tools could be used; we use the results from [204] primarily because they are amenable to real-time computation [214].…”

Section: Ergodic Controlmentioning

confidence: 99%

“…Distributed data collection of this kind has been widely and successfully applied in a variety of contexts, such as environmental monitoring [236,269]. The key feature underlying the success of these distributed control applications is that the dynamics of the robot collective are factorable into a block-diagonal representation-the dynamics of each robot agent are independent from one another [214,270]. However, can we expect this to be the case across active learning applications?…”

Section: Distributabilitymentioning

confidence: 99%

See 1 more Smart Citation

Active Learning in Robotics: A Review of Control Principles

Taylor,

Berrueta,

Murphey

2021

Preprint

View full text Add to dashboard Cite

Active learning is a decision-making process. In both abstract and physical settings, active learning demands both analysis and action. This is a review of active learning in robotics, focusing on methods amenable to the demands of embodied learning systems. Robots must be able to learn efficiently and flexibly through continuous online deployment. This poses a distinct set of control-oriented challenges-one must choose suitable measures as objectives, synthesize realtime control, and produce analyses that guarantee performance and safety with limited knowledge of the environment or robot itself. In this work, we survey the fundamental components of robotic active learning systems. We discuss classes of learning tasks that robots typically encounter, measures with which they gauge the information content of observations, and algorithms for generating action plans. Moreover, we provide a variety of examples-from environmental mapping to nonparametric shape estimation-that highlight the qualitative differences between learning tasks, information measures, and control techniques. We conclude with a discussion of control-oriented open challenges, including safety-constrained learning and distributed learning.

show abstract

Section: Ergodic Controlmentioning

confidence: 99%

Section: Distributabilitymentioning

confidence: 99%

Active Learning in Robotics: A Review of Control Principles

Taylor,

Berrueta,

Murphey

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…For example, authors in [9] developed a system that enabled human operators to control their swarm through density specifications via a touchscreen interface. Authors in [10] created an end-to-end swarm control system that leveraged their swarm's heterogeneous capability, used autonomously detected information to keep the swarm safe, and enabled operators to specify multimodal commands to their swarm via touchscreen that were adaptable in real-time. Furthermore, the touchscreen interfaces used in these works to send commands enabled both persistent swarm behavior and multiple variations of operator commands to achieve tasks through density specifications.…”

Section: Introductionmentioning

confidence: 99%

A Game Benchmark for Real-Time Human-Swarm Control

Meyer¹,

Pinosky²,

Thomas³

et al. 2022

2022 IEEE 18th International Conference on Automation Science and Engineering (CASE)

View full text Add to dashboard Cite

We present a game benchmark for testing humanswarm control algorithms and interfaces in a real-time, highcadence scenario. Our benchmark consists of a swarm vs. swarm game in a virtual ROS environment in which the goal of the game is to "capture" all agents from the opposing swarm; the game's high-cadence is a result of the capture rules, which cause agent team sizes to fluctuate rapidly. These rules require players to consider both the number of agents currently at their disposal and the behavior of their opponent's swarm when they plan actions. We demonstrate our game benchmark with a default human-swarm control system that enables a player to interact with their swarm through a high-level touchscreen interface. The touchscreen interface transforms player gestures into swarm control commands via a low-level decentralized ergodic control framework. We compare our default humanswarm control system to a flocking-based control system, and discuss traits that are crucial for swarm control algorithms and interfaces operating in real-time, high-cadence scenarios like our game benchmark. Our game benchmark code is available on Github; more information can be found at https: //sites.google.com/view/swarm-game-benchmark.

show abstract

“…In this paper, we present a multi-agent decentralized trajectory planning problem in the framework of ergodic exploration. Related works are: [11], which reformulates the singleagent problem in [4] using a Nash equillibrium interpretation for the multi-agent setting; [12], focusing on area coverage with obstacles; and [13] which demonstrates a decentralized ergodic swarm control framework adaptable to external user commands and dynamic environmental information.Ṫhe contribution of this work is twofold. First, a decentralized multiagent extension of the approach in [4] is proposed (Section III).…”

Section: Introductionmentioning

confidence: 99%

Decentralized Trajectory Optimization for Multi-Agent Ergodic Exploration

Gkouletsos

Iannelli

Badyn

et al. 2021

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

Autonomous exploration is an application of growing importance in robotics. A promising strategy is ergodic trajectory planning, whereby an agent spends in each area a fraction of time which is proportional to its probability information density function. In this paper, a decentralized ergodic multi-agent trajectory planning algorithm featuring limited communication constraints is proposed. The agents' trajectories are designed by optimizing a weighted cost encompassing ergodicity, control energy and close-distance operation objectives. To solve the underlying optimal control problem, a second-order descent iterative method coupled with a projection operator in the form of an optimal feedback controller is used. Exhaustive numerical analyses show that the multi-agent solution allows a much more efficient exploration in terms of completion task time and control energy distribution by leveraging collaboration among agents.

show abstract

Ergodic Specifications for Flexible Swarm Control: From User Commands to Persistent Adaptation

Cited by 20 publications

References 16 publications

Active Learning in Robotics: A Review of Control Principles

Active Learning in Robotics: A Review of Control Principles

A Game Benchmark for Real-Time Human-Swarm Control

Decentralized Trajectory Optimization for Multi-Agent Ergodic Exploration

Contact Info

Product

Resources

About