A5052225367- Modular Robots and Swarm Intelligence
- Evolutionary Game Theory and Cooperation
- Complex Network Analysis Techniques
- Distributed Control Multi-Agent Systems
- Opinion Dynamics and Social Influence
- Insect and Arachnid Ecology and Behavior
- Gamma-ray bursts and supernovae
- Blockchain Technology Applications and Security
- Distributed systems and fault tolerance
- Reinforcement Learning in Robotics
- Plant and animal studies
- Neural dynamics and brain function
- Opportunistic and Delay-Tolerant Networks
- Robotics and Sensor-Based Localization
- Energy Efficient Wireless Sensor Networks
- Neural and Behavioral Psychology Studies
- Slime Mold and Myxomycetes Research
- Optimization and Search Problems
- SAS software applications and methods
- Robotic Path Planning Algorithms
- Animal Behavior and Reproduction
- Decision-Making and Behavioral Economics
- Neurobiology and Insect Physiology Research
- IoT and Edge/Fog Computing
- Robotic Locomotion and Control
Université Libre de Bruxelles
2014-2024
University of Konstanz
2023-2024
Max Planck Institute of Animal Behavior
2024
University of Sheffield
2016-2023
Hospital General Universitario de Elche
2013
Dalle Molle Institute for Artificial Intelligence Research
2011
The engineering of large-scale decentralised systems requires sound methodologies to guarantee the attainment desired macroscopic system-level behaviour given microscopic individual-level implementation. While a general-purpose methodology is currently out reach, specific solutions can be broad classes problems by means well-conceived design patterns. We propose pattern for collective decision making grounded on experimental/theoretical studies nest-site selection observed in honeybee swarms...
To effectively perform collective monitoring of dynamic environments, a robot swarm needs to adapt changes by processing the latest information and discarding outdated beliefs. We show that in composed robots relying on local sensing, adaptation is better achieved if have shorter rather than longer communication range. This result contrast with widespread belief more links always improve exchange network. tasked reaching agreement best option currently available their operating environment....
This paper presents a model for collective decision making applied to swarm of honeybees choosing nest. The authors tackle the extension known binary option an arbitrary number choices, and study situation when one is better than rest.
Working with large swarms of robots has challenges in calibration, sensing, tracking, and control due to the associated scalability time requirements. Kilobots solve this through their ease maintenance programming, are widely used several research laboratories worldwide where low cost enables large-scale studies. However, small, inexpensive nature limits range capabilities as they only equipped a single sensor. In some studies, limitation can be source motivation inspiration, while others it...
Abstract How groups of cooperative foragers can achieve efficient and robust collective foraging is interest both to biologists studying social insects engineers designing swarm robotics systems. Of particular are distance-quality trade-offs swarm-size-dependent strategies. Here, we present a system based on virtual pheromones, tested in simulation swarms up 200 physical robots. Our individual agent controllers highly simplified, as they binary pheromone sensors. Despite being simple, our...
Abstract Through theoretical analysis, we show how a superorganism may react to stimulus variations according psychophysical laws observed in humans and other animals. We investigate an empirically-motivated honeybee house-hunting model, which describes value-sensitive decision process over potential nest-sites, at the level of colony. In this study, colony time increases with number available nests, agreement Hick-Hyman law psychophysics, decreases mean nest quality, Piéron’s law. also that...
Abstract Strongly opinionated minorities can have a dramatic impact on the opinion dynamics of large population. Two factions inflexible minorities, polarised into two competing opinions, could lead entire population to persistent indecision. Equivalently, populations remain undecided when individuals sporadically change their based individual information rather than social information. Our analysis compares cross-inhibition model with voter for decisions between equally good alternatives,...
To succeed in their objectives, groups of individuals must be able to make quick and accurate collective decisions on the best option among a set alternatives with different qualities. Group-living animals aim do that all time. Plants fungi are thought so too. Swarms autonomous robots can also programed best-of-n for solving tasks collaboratively. Ultimately, humans critically need it many times they should better at it! Thanks mathematical tractability, simple models like voter model local...
We present a novel technology that allows real robots to perceive an augmented reality environment through virtual sensors. Virtual sensors are useful and desirable for research activities because they allow researchers quickly efficiently perform experiments would otherwise be more expensive, or even impossible. In particular, is (i) prototyping assessing the impact of new before physically produced; (ii) developing studying behaviour should deal with phenomena cannot easily reproduced in...
Consensus dynamics in decentralised multiagent systems are subject to intense studies, and several different models have been proposed analyzed. Among these, the naming game stands out for its simplicity applicability a wide range of phenomena applications, from semiotics engineering. Despite studies available, implementation theoretical real distributed is not always straightforward, as physical platform imposes constraints that may bearing on consensus dynamics. In this letter, we...
It is usually assumed that information cascades are most likely to occur when an early but incorrect opinion spreads through the group. Here, we analyse models of confidence-sharing in groups and reveal opposite result: simple plausible naive-Bayesian decision-making exhibit group decisions synchronous; however, asynchronous, reached by Bayesian decision-makers tend be correct dominate consensus dynamics. Thus actually rescue from making errors, rather than contribute it. We explore realism...
We investigate decentralised decision-making, in which a robot swarm is tasked with selecting the best-quality option among set of alternatives. Individual robots are simplistic as they only perform diffusive search, make local noisy estimates options' quality, and exchange information near neighbours. propose algorithm, inspired by house-hunting honeybees, to efficiently aggregate estimations. robots, varying over time single parameter that modulates interaction strength, balance...
Efficient engineered systems require scalability. A scalable system has increasing performance with size. In an ideal situation, the increase in (e.g., speedup) corresponds to number of units processors, robots, users) that are added three times processors a computer would lead faster computations). However, if multiple work on same task, then coordination among these is required. This can introduce overheads impact performance. The costs sublinear improvement or even diminishing there also...
Robot swarms are generally considered to be composed of cooperative agents that, despite their limited individual capabilities, can perform difficult tasks by working together. However, in open swarms, where different robots added the swarm parties with potentially competing interests, cooperation is but one many strategies. We envision an information market buy and sell through transactions stored on a distributed blockchain, encouraged economy itself. As proof concept, we study classical...