Distributed Network Sensing as an Escape Strategy

Swarm behavior in both biological and robotic systems poses a significant challenge due to the need for decentralized decision-making. A potential solution is embodied distributed sensing and decision-making, where local environmental cues lead to coordinated changes in the system. Fluidic networks provide a physical foundation for such systems, although current implementations are pre-designed and lack the ability to process spatial information across large systems.

In this study, researchers examine how fluid transport networks in the slime mould Physarum polycephalum enable environmental sensing and adaptive responses. By exposing Physarum's vein networks to varying pH conditions, the study reveals that the organism's ability to respond to unpleasant stimuli depends on its size and network connectivity.

Organisms larger than a certain size can reliably detect and escape harmful conditions by reconfiguring their network structure. However, disconnected or spatially confined organisms cannot respond effectively. In environments with conflicting stimuli, the network's morphology allows for the integration of spatial information and repositioning into more favorable areas.

Local contractions within the network serve as the physical mechanism for redistributing fluids and converting localized stimulation into system-wide responses. The researchers generate various morphologies using fluidic and electrochemical control, demonstrating the ability to adapt large-scale systems to complex environments. These findings establish distributed fluidic networks as a framework for integrated sensing and actuation, offering design principles for autonomous soft robotic systems that operate in dynamic settings.