Vendredi 28 novembre 2025 à 11h00, salles des séminaires IRPHE
Abstract: A central goal in biorobotics and tissue engineering is to develop smart, adaptive systems capable of perceiving and responding to their environment. Achieving such autonomy across multiple scales requires not only local sensing and actuation, but also mechanisms for distributing materials and signals over long distances. Biological vascular networks provide a compelling model for this, as they enable long-range transport, environmental responsiveness, and morphological reconfiguration: key features of scalable, adaptive systems.
In this work, we explore how vascular-like functionality can be integrated into bioelectronic interfaces using the unicellular slime mold Physarum polycephalum. This organism forms a dynamic, self-organized tubular network that exhibits emergent behaviors such as optimization, adaptation, and self-healing, making it an ideal model for studying decentralized, stimulus-responsive growth. By leveraging electrochemically induced pH gradients, we experimentally translate designed network layouts into biological structures and characterize the key dynamics underlying their responses to pH stimuli.
We further analyze Physarum’s morphological evolution, where the network reorganizes over several hours to evacuate a defined area by sequentially pruning competing parallel veins, ultimately forming a tree-like architecture. Drawing an analogy with power-grid networks, we investigate how sequential pruning depends on the ratio of tube to network resistance. Analytical and numerical results show that regular graphs undergo pruning until the average node degree falls below four, a finding that remains robust in simulations of random networks. Incorporating mass redistribution into this process leads to resistance homogenization, revealing fundamental physical constraints that shape adaptive transport networks.
Mathieu le Verge-Serandour, School of Natural Sciences, Technical University of Munich