Dresden 2026 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

SOE: Fachverband Physik sozio-ökonomischer Systeme

SOE 4: Poster Session

SOE 4.1: Poster

Monday, March 9, 2026, 18:00–21:00, P4

Artificial "Minimal" Intelligence in a Decentralized Mechanical System — •Alex Leffell and Manu Prakash — Stanford University, California, USA

Computation is not exclusive to neurons or transistors. Simple, aneural organisms like Placozoa exhibit complex behaviors, demonstrating that intelligence can arise from non-neuronal physical dynamics. We model these organisms as Phased Active Elastic Sheets (PhAES) – networks of coupled active oscillators that compute via the dynamics of their vibrational modes. This approach builds on established work in both physical learning and active solids and extends it to understand animal behavior. Using both in-silico and robophysical models, we show that this system can perform complex behaviors like gradient climbing without any implicit memory and purely mechanical communication between cells. This perspective offers a direct path towards a new class of computational materials that physically embody sensing, actuation, and learning. Such systems would bridge the gap between rigid, centrally-controlled robots and task-specific soft matter, fundamentally expanding our notions on the physical basis of intelligence.

Keywords: physical learning; embodied cognition; morphological computation