A reprogrammable chip learns to read light's hidden knots

Some patterns in physics are stubborn: knot-like features in waves that survive defects and noise, called topological properties, are prized for building more reliable devices — but they are hard to detect in real systems, which leak energy. At KTH Royal Institute of Technology in Sweden, a team led by Ali Elshaari, with lead author Andrea Cataldo, has demonstrated a programmable photonic chip — light-based circuitry — that can be reconfigured to run many tests on one device and reveal those hidden features, even pinpointing an 'exceptional point' where a system's behaviour flips. Published in Physical Review Letters, the method could speed the search for sturdier optical, electronic and quantum technologies.

SDG 9 · Industry, innovation and infrastructure — Mechanism: tech innovation