Article
Computer Science, Artificial Intelligence
Jian Yang, Yuhui Shi
Summary: With the increasing complexity of tasks and uncertainty in the environment, achieving adaptability and robustness for multi-robot cooperation tasks through manual design methods is challenging. Automatic synthesis approaches with trial and error mechanisms are gaining more attention. The proposed Brain Storm Robotics (BSR) framework encodes strategies as "ideas" and can obtain satisfactory solutions for specific tasks through a series of operations on these ideas. This paper proposes an automatic design approach for neural network-based strategies using the BSR framework to achieve cooperative behaviors in robotic swarms.
APPLIED SOFT COMPUTING
(2022)
Article
Robotics
Martin Jilek, Michael Somr, Miroslav Kulich, Jan Zeman, Libor Preucil
Summary: The research introduces a framework for designing macroscale passive robots capable of targeted self-assembly, utilizing properly designed magneto-mechanical locks for jamming-free assembly, dedicated encoding of glues for guiding tile interactions, and consistent formalization of geometrical constraints to ensure valid assembly.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Computer Science, Information Systems
Yuan Xu, Gelei Deng, Tianwei Zhang, Han Qiu, Yungang Bao
Summary: The development of robotics technology is accelerated by strong support from cloud computing, but also faces potential DoS threats. This paper introduces new DoS attack methods and demonstrates their severity through evaluations and case studies, urging the robotics community to enhance security measures and performance in cloud-robotic systems.
INFORMATION SCIENCES
(2021)
Article
Automation & Control Systems
Vu Phi Tran, Matthew A. Garratt, Kathryn Kasmarik, Sreenatha G. Anavatti
Summary: This paper proposes a novel swarm-based control algorithm for multi-robot exploration and repeated coverage in environments with unknown, dynamic obstacles. The algorithm combines frontier-led swarming for driving exploration and pheromone-based stigmergy for controlling repeated coverage while avoiding obstacles. Through comparison experiments, it is demonstrated that the proposed strategy outperforms other existing multi-robot repeated coverage methodologies.
IEEE-CAA JOURNAL OF AUTOMATICA SINICA
(2023)
Article
Robotics
Yulun Tian, Yun Chang, Fernando Herrera Arias, Carlos Nieto-Granda, Jonathan P. How, Luca Carlone
Summary: This paper presents $\mathsf {\text{Kimera-Multi}}$, a multi-robot SLAM system that is robust, fully distributed, and capable of capturing semantic information. Experimental results demonstrate its superior performance.
IEEE TRANSACTIONS ON ROBOTICS
(2022)
Article
Robotics
Shuai Zhang, Jia Pan
Summary: This letter presents a distributed approach for automatically collecting a flock of robots. A density-based strategy is proposed to address the challenge of collecting a larger flock with a limited number of robots. The approach combines edge-following behavior and a shrink mechanism to successfully collect both single clusters and flocks with multiple sub-groups.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Chemistry, Multidisciplinary
Alexandre Campo, Stamatios C. Nicolis, Jean-Louis Deneubourg
Summary: Remembering information is crucial for adaptation and behavior adjustment in both living organisms and artificial systems. Pavlov's experiments demonstrate how single animals can trigger behavioral responses based on memory of stimuli associated with rewards. Researchers propose a novel behavior based on aggregation process to enable robotic swarm to exhibit collective memory.
APPLIED SCIENCES-BASEL
(2021)
Article
Computer Science, Information Systems
Sanghyeon Bae, Sunghyeon Joo, Junhyeon Choi, Jungwon Pyo, Hyunjin Park, Taeyong Kuc
Summary: Multi-robot systems face the challenge of simultaneously considering robot movement and influence while planning tasks. This paper proposes a semantic knowledge-based hierarchical planning approach that extends knowledge by considering interactions between environmental elements. The task planner utilizes spatial hierarchy knowledge to group robots and generate optimal task plans for each group.
Article
Computer Science, Artificial Intelligence
David M. Bossens, Danesh Tarapore
Summary: This article introduces an algorithm called QED, which enhances the quality of evolved archives by inducing behavioral diversity through introducing environmental diversity, leading to improved recovery capability for robot swarms when facing faults. The study shows that the archives evolved by QED outperform traditional behavioral descriptors in 5 different robot swarm benchmark tasks.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2021)
Article
Computer Science, Artificial Intelligence
Melanie Schranz, Gianni A. Di Caro, Thomas Schmickl, Wilfried Elmenreich, Farshad Arvin, Ahmet Sekercioglu, Micha Sende
Summary: Swarm Intelligence is a multi-agent framework inspired by natural systems, where each agent acts autonomously and collaborates with others without central control. While well-designed swarms have advantages in adaptability, robustness, and scalability, they have not yet transitioned from lab demonstrations to real-world applications, especially in embodied scenarios like swarm robotics.
SWARM AND EVOLUTIONARY COMPUTATION
(2021)
Article
Computer Science, Artificial Intelligence
Antoine Ligot, Andres Cotorruelo, Emanuele Garone, Mauro Birattari
Summary: This article proposes an experimental protocol for comparing fully automatic design methods for robot swarms, including defining benchmarks and sampling strategies, to address the lack of systematic analysis and comparison in the optimization-based design of robot swarms.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2022)
Article
Robotics
Changrak Choi, Muhammad Adil, Amir Rahmani, Ramtin Madani
Summary: Multi-robot systems have enhanced capabilities but increased coordination complexity. To address this, we propose a convexification method, parabolic relaxation, to generate feasible trajectories. Numerical experiments demonstrate high success rate and feasibility.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Mathematics, Applied
Qi Zhu, Weihua Yang, Xiaohong Zhang, Ben Gao, Jinwei Yu
Summary: This paper explores the region reaching consistency for uncertain fully-actuated multi-robot systems and proposes a distributed region reaching consensus control strategy, discussing its characteristics and scope.
ZAMM-ZEITSCHRIFT FUR ANGEWANDTE MATHEMATIK UND MECHANIK
(2023)
Article
Computer Science, Artificial Intelligence
Paulo Rezeck, Hector Azpurua, Mauricio F. S. Correa, Luiz Chaimowicz
Summary: This paper presents a novel platform design for swarm robotics applications that is low cost and easy to assemble. The platform is deeply integrated with the widely used ROS framework and consists of a 3D printed body and open-source software. The functionality of the platform is evaluated through experiments, demonstrating its capability in performing different swarm tasks despite its small size and reduced cost, making it suitable for swarm robotics research and education.
Article
Computer Science, Artificial Intelligence
Fernando J. Mendiburu, David Garzon Ramos, Marcos R. A. Morais, Antonio M. N. Lima, Mauro Birattaria
Summary: This paper presents Mate, an automatic off-line design method for spatially-organizing behaviors in robot swarms. Experimental results show that Mate outperforms other automatic design methods in handling tasks with spatial distribution constraints.
SWARM AND EVOLUTIONARY COMPUTATION
(2022)
Article
Thermodynamics
Benoit Delcroix, Jerome Le Ny, Michel Bernier, Muhammad Azam, Bingrui Qu, Jean-Simon Venne
Summary: Thermal models of buildings are essential for forecasting energy use and improving mechanical system control. Autoregressive neural network models outperform gray-box and black-box linear models in simulating indoor temperatures with limited information.
BUILDING SIMULATION
(2021)
Article
Robotics
Mohammed Shalaby, Charles Champagne Cossette, Jerome Le Ny, James Richard Forbes
Summary: This paper introduces a novel cascaded and decentralized filtering approach that approximates cross-covariances when local estimators consider distinct state vectors. The proposed estimator is validated in simulations and experiments on a three-dimensional attitude and position estimation problem, showing superior performance compared to other filtering approaches.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Robotics
Charles Champagne Cossette, Mohammed Shalaby, David Saussie, James Richard Forbes, Jerome Le Ny
Summary: The paper introduces an algorithm for determining the three-dimensional relative position between two mobile robots using inexpensive sensors. The algorithm has been tested in simulation and experiment, outperforming standard estimators and suitable for real-time implementation.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Robotics
Daniil Lisus, Charles Champagne Cossette, Mohammed Shalaby, James Richard Forbes
Summary: This letter demonstrates how to estimate robot heading using UWB range and RSS measurements, by learning a data-driven relationship and combining with a gyroscope and an invariant extended Kalman filter.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Robotics
Justin Cano, Gael Pages, Eric Chaumette, Jerome LeNy
Summary: Ultra-Wide Band (UWB) communication systems can be used to design low cost and power-efficient navigation systems for mobile robots. However, even in favorable conditions, significant errors are often observed in the Time of Flight (ToF) measurements. This letter proposes a ToF error model and a bias compensation scheme using on-board measurements to achieve accurate ranging accuracy.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Automation & Control Systems
Kwassi H. Degue, Jerome Le Ny
Summary: When multiple agents solve a joint optimal control problem cooperatively, coordinating their control signals is beneficial, but sharing local measurements might compromise privacy. This study addresses the Linear Quadratic Gaussian (LQG) control problem with differential privacy constraints, ensuring the published signals are not too sensitive to any single agent's data. A two-stage architecture for differentially private LQG control is proposed, leveraging a previously developed Kalman filtering solution. The architecture's first stage can be implemented by a coordinator aggregating and perturbing measurements or without a trusted aggregator using a secure sum protocol. Numerical simulations demonstrate the performance improvement over simpler alternatives like directly perturbing agent measurements.
IEEE CONTROL SYSTEMS LETTERS
(2023)
Article
Computer Science, Artificial Intelligence
Tzu-Yi Chiu, Jerome Le Ny, Jean-Pierre David
Summary: For complex automated perception and decision tasks, it is important to develop trust and understand the behavior of algorithms that may be too complex for human users. This article combines the anchors methodology with Monte Carlo Tree Search to provide explanations for black-box models' decisions. The methodology searches for descriptive explanations in the form of input signal properties, expressed in Signal Temporal Logic, to reproduce observed behavior.
ARTIFICIAL INTELLIGENCE
(2023)
Article
Robotics
Justin Cano, Jerome Le Ny
Summary: In robotic networks relying on noisy range measurements between agents for cooperative localization, potential-based planning methods are introduced to characterize the quality of the network geometry for cooperative position estimation. These methods utilize Cramer Rao lower bounds (CRLB) to provide a theoretical lower bound on the achievable positioning accuracy. The concept of localizability is extended using equality-constrained CRLBs to scenarios where additional information about the relative positions of the ranging sensors is known.
IEEE TRANSACTIONS ON ROBOTICS
(2023)
Proceedings Paper
Automation & Control Systems
Kwassi H. Degue, Karthik Gopalakrishnan, Max Z. Li, Hamsa Balakrishnan, Jerome Le Ny
Summary: This paper explores the detection of outliers in data while preserving individual privacy. The algorithm utilizes sparse vector technique in multivariate Gaussian signals and quantifies the trade-off between accuracy and privacy. The analytical results are validated through numerical simulations.
2021 AMERICAN CONTROL CONFERENCE (ACC)
(2021)
