Team
CAB
Team manager: Fourcassie Vincent & Jost Christian
Presentation
The main objective of our team is to understand the behavioral and cognitive mechanisms underlying collective behaviors and collective decisions in groups of animals. Our aim is to identify the mechanisms that allow a group of animals to coordinate their actions and to provide adaptive collective responses to the changes occurring in its environment, i.e. to display collective intelligence. We try to understand in particular how the ability of a group to collectively solve problems varies with the properties of the individuals it is composed of. The general methodology of our work is based on a tight combination of experimental and modeling approaches.
Project 1
This project aims on the one hand to characterize and model the functional form of social interactions involved in the coordination of 3D movements and the transmission of information in groups of fish and, on the other hand, to analyze avoidance strategies and collective responses of these groups during a controlled perturbation simulating a predator attack. The comparative study will be carried out on three species of fish Hemigrammus rhodostomus, Oryzias latipes and Puntigrus tetrazona which exhibit pronounced schooling swimming behavior but different swimming speeds and different forms of collective movements. The analysis, the modeling of interactions between fish and finally the validation of the model will be carried out using the closed loop virtual reality device developed at the CRCA in collaboration with IRIT.
Project 2
This project aims to study the collective behavior of fish and their social interactions under stressful conditions and how, through the modulation of these interactions, stress can propagate within a school and induce changes in collective phases. In this study we will use a closed-loop virtual reality setup to analyze in real time the interactions between a real fish and a virtual fish whose behavior is controlled by a model.
Project 3
In this project we will study under what conditions a group of pedestrians can collectively optimize their choices in a minority game in a virtual environment. The experiments will be carried out with virtual reality headsets. The objective is to determine the most relevant types of information to deliver to a group of pedestrians in order to collectively optimize the group performance in a virtual minority game. The study will also involve a comparison of the walking of humans in a virtual environment with that in a real physical environment, by measuring avoidance interactions in these two contexts using the methods that we have introduced in the study of interactions between fish in a school, and by exploiting the human trajectory monitoring methods developed by the LAAS.
Project 4
This project aims to develop an autonomous swarm of drones based on bio-inspired principles. The key objectives are to design locally optimized control mechanisms guiding the collective motion and to provide theoretical measures for real-time characterization of swarm cohesion while maneuvering in uncertain and complex dynamic environments.
Project 5
We aim at studying the development of the locomotion in juveniles and to what extent they modify their own locomotion behaviour to cope with adults’ behaviour. We use sheep as a model to answer this question. This species presents the particularity to alternate stationary phases when grazing and collective movement phases. Our main hypothesis is that juveniles suffer from an energetic cost when following adults, due to their social attraction. In our experiments, we monitor groups containing only juveniles, only adults and groups containing as much as adults as juveniles to study the locomotion and socio-spatial cohesion. Locomotion is obtained using high-precision GNSS.
Project 6
In summer, peaks of high temperature are more and more common. During these periods, the temperatures are largely above the neutral thermic zone of animals. During these periods of thermic stress, sheep commonly aggregate in highly dense groups, which is counter-intuitive. We aim at studying the dynamics of flocking during periods outside and during thermic stress periods. If aggregation is probably beneficial, we also suspect that it generates detrimental local atmospheric condition. We intend equipping animals with different sensors to measure physiological responses as well as environmental conditions. We will compare the behaviour of animals in pasture with and without solar screens.
Project 7
With about 16,000 species identified to date, ants are a family of insects presenting a remarkable biodiversity. They are also found in a wide variety of ecological environments. The aim of this project is to characterize their locomotion from a biomechanical point of view and to identify the structural and behavioral adaptations that allow them to move at very high speeds over a wide variety of substrates, or to lift and carry loads weighing more than ten times their own weight. To this end, we use a platform consisting of five synchronized cameras to study the kinematics of their locomotion, and a captor developed in collaboration with the LAAS to measure the ground reaction forces exerted on their legs. Ultimately, our aim is to establish a link between the biomechanics of ant locomotion and their performances in load transport, whether this latter is achieved individually or collectively, when several ants cooperate to move an object.
Project 8
The natural environment of ants and termites is complex, requiring movement on 2D surfaces within a 3D space. This movement influences collective behaviors such as exploration, foraging, and nest construction. Our project focuses on observing and modeling insect movement on controlled curved surfaces, which may be combined with pheromone trail laying to create distinct movement trails that increase efficiency. To assess this efficiency, we employ advanced computer simulations in collaboration with physicists, using null collision models to study the benefits and limitations of these modern numerical schemes in understanding collective animal behaviors.
Project 9
The project aims to develop innovative methodologies for analyzing collective behavior. We focus our interest on team sports, and especially rugby, with a focus on modeling and predicting performance. We view team performance as a correlated time series prediction task, where the goal is to forecast the success of a team based on the analysis of player movement and collective dynamics. Recent studies have shown that traditional methods for analyzing team sports are limited, and a paradigm shift is needed to better understand the complex interactions between players. Our project addresses this challenge by combining statistical modeling, machine learning, and data analysis to develop predictive models of group dynamic and team performance. To tackle this question we use relevant data including GPS tracking, accelerometers, and video recordings, to extract relevant features that describe player behavior and team dynamics. We search for patterns and correlations in player movements, such as synchronization, coordination, and decision-making processes. Our project has the potential to bring new tools to the analysis of team sports, providing coaches, trainers, and analysts with data-driven insights to improve team performance. By modeling and predicting success as a correlated time series prediction task, we aim to uncover the underlying mechanisms that drive team success, and to develop predictive models that can inform decision-making in real-world sports settings.
Project 10
The ‘Dimensionless Approach’ project delves into the intersection of biomechanics, physics, and biomimetics to uncover new principles of locomotion and extraordinary efficiency observed in insects. By leveraging the concept of dimensionless analysis, this multidisciplinary initiative examines space-time dilation effects to better understand the behaviors of “super-efficient” insects and their potential technological applications. It integrates three subprograms:
– Sub-program 1 investigates the interplay between ants and their environment focusing on how the chemical composition of their cuticle impacts physical interactions across environment and individuals leading to a biomechanical efficiency.
– Sub-program 2 integrates a more statistical physics methodologies to study the collective load transport in ants. This research unravels the dynamics of group cooperation and movement, offering insights into the mechanics of coordinated activity in complex systems.
– Sub-program 3 channels these biological findings into biomimetic applications. It aims to design exoskeletons inspired by the extraordinary capabilities of ants, develop robotic hexapod legs with advanced mobility, and create collective load-transport assistive devices for human use.
This project, positioned at the convergence of biological discovery and engineering ingenuity, exemplifies the transformative potential of a dimensionless approach in bridging natural and artificial systems.
Team members
– Bassanetti, T., Escobedo, R., Cezera, S., Blanchet, A., Sire, C. & Theraulaz, G. 2023. Cooperation and deception through stigmergic interactions in human groups. Proceedings of The National Academy of Sciences USA, 120 (42), e2307880120.
– Dorigo, M., Theraulaz, G. & Trianni, V. 2021. Swarm robotics: past, present and future. Proceedings of the IEEE, 109: 1152-1165.
– Escobedo, R., Lecheval, V., Papaspyros, V., Bonnet, F., Mondada, F., Sire, C., & Theraulaz, G. 2020. A data-driven method for reconstructing and modelling social interactions in animal groups. Philosophical Transactions of the Royal Society of London – Serie B., 375, 20190380.
– Gómez-Nava, L., Bon, R., Peruani, F. 2022. Intermittent collective motion in sheep results from alternating the role of leader and follower. Nature Physics, 18, 1494-1501.
– Heyde A, Guo L, Jost C, Theraulaz G, Mahadevan L. 2021. Self-organized Biotectonics of termite nests. Proceedings of The National Academy of Sciences USA, 118 (5), e2006985118, doi: 10.1073/pnas.2006985118
– Lei, L., Escobedo, R., Sire, C., Theraulaz, G. 2020. Computational and robotic modeling reveal parsimonious combinations of interactions between individuals in schooling fish. Plos Computational Biology, 16: e1007194.
– Merienne, H., Latil, G., Moretto, P., Fourcassié, V. 2020. Walking kinematics in the polymorphic seed harvester ant Messor barbarus: influence of body size and load carriage. Journal of Experimental Biology, 223, jeb.205690.
– Xue, T., Li, X., Lin, G., Escobedo, R., Sire, C., Han, Z., Chen, X. & Theraulaz, G. 2023. Tuning social interactions’ strength drives collective response to light intensity in schooling fish Plos Computational Biology, 19(11):e1011636.
Affiliation
