Keywords: Swarm Robotics, Spectral Analysis, Optimization, Information Diffusion, Symbolic Regression, Swarm Intelligence, Quality-Diversity algorithms
Abstract:
Swarm robotics [1] investigates decentralized, collaborative behaviors in large groups of simple robots inspired by natural systems like ants or bees. In this field, the diffusion of local information among agents is critical for enabling collective decision-making and adaptation to tasks and environments. However, traditional approaches largely rely on hand-crafted microscopic rules [2], which limit scalability and adaptability for complex applications.
This project aims to advance swarm robotics by automating the discovery of spectral diffusion operators [3] that enable robust information propagation within robotic swarms. Building on our preliminary results [4], where Kilobots robots [5] were programmed to emulate the Laplacian operator to achieve collective shape classification (i.e. identify disk- or ring-shaped arenas), we hypothesize that more sophisticated diffusion schemes can address increasingly complex tasks. These schemes involve nonlinear dynamics and high-dimensional optimization problems, making manual design difficult.
To overcome these challenges, we propose using artificial intelligence techniques, including (physics-informed) Artificial Neural Networks, Evolutionary Algorithms [7] and Quality-Diversity algorithms [8], to discover novel diffusion operators. These operators could extend beyond traditional paradigms to include reaction-diffusion systems or entirely new modes of information sharing, such as positional or gradient-based schemes.
Our findings aim to establish the emerging field of spectral swarm robotics as a foundation for tackling advanced applications like collective sensing, decision-making, and task allocation. This work represents a transformative step toward extending swarm robotics methods to complex tasks and real-world scenarios.
Internship terms and conditions:
This 5/6-month interdisciplinary M2 internship is at a crossroad between swarm robotics, artificial intelligence, and mathematical models of information diffusion. It will take place at the ISIR lab (Institut des Systèmes Intelligents et de Robotique) at Sorbonne Université, among an interdisciplinary team composed of computer scientists, roboticists and biophysicists. It will be funded by the Spectral-Swarm-Robotics of the ANR, with the opportunity to continue as a PhD Student for promising candidates.
It will be jointly supervised by Dr Leo Cazenille and Pr Nicolas Bredeche (ISIR, SU).
Potential publication opportunities in AI/robotics conferences.
Expected skills and research interests:
- Master 2 level in AI / Computer Science, Bio-physics or Applied Mathematics
- Proficiency in the Python language
- Optional knowledge of data-science, machine learning libraries, such as pandas, pytorch, etc..
Réferences:
[1] H Hamann. Swarm robotics: A formal approach, volume 221. 2018.
[2] M Dorigo, G Theraulaz, and V Trianni. Reflections on the future of swarm robotics. Science Robotics, 2020.
[3] D Spielman. Spectral graph theory. Combinatorial scientific computing, 2012.
[4] Cazenille, L, N Lobato-Dauzier, A Loi, M Ito, O Marchal, Aubert-Kato, N, Bredeche, N, and Genot, AJ. Hearing the shape of an arena with spectral swarm robotics. arXiv:2403.17147, 2024.
[5] M Rubenstein, A Cornejo, and R Nagpal. Programmable self-assembly in a thousand-robot swarm. Science, 2014.
[6] Raissi, Maziar, Paris Perdikaris, and George E. Karniadakis. "Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations." Journal of Computational physics 378 (2019): 686-707.
[7] Hansen, Nikolaus, Sibylle D. Müller, and Petros Koumoutsakos. "Reducing the time complexity of the derandomized evolution strategy with covariance matrix adaptation (CMA-ES)." Evolutionary computation 11.1 (2003): 1-18.
[8] Chatzilygeroudis, Konstantinos, et al. "Quality-diversity optimization: a novel branch of stochastic optimization." Black Box Optimization, Machine Learning, and No-Free Lunch Theorems. Cham: Springer International Publishing, 2021. 109-135.