Researchers Demonstrate Fiber Sensor With <3 ns Readout and 100% Torsion Classification Accuracy
Updated
Updated · Nature.com · May 25
Researchers Demonstrate Fiber Sensor With <3 ns Readout and 100% Torsion Classification Accuracy
1 articles · Updated · Nature.com · May 25
<3 ns demodulation delay was achieved in a new all-optical fiber sensing architecture, AOFS-IC, which reads physical changes directly from light intensity without electronic signal processing.
The system combines a scattering medium with a diffractive optical network to map perturbations to output intensity at light speed, cutting latency by more than two orders of magnitude versus conventional fiber sensing.
Tests showed sub-nano strain resolution, including 1.6160 nε RMSE over a 95 nε range, and 100% classification accuracy for nine torsion states spaced 45° apart.
AOFS-IC also multiplexed strain and torsion sensing from one multimode fiber and monitored a 3-DOF robotic arm, with joint-angle RMSEs of 1.7071°, 1.7003° and 1.8755°.
The researchers say replacing programmable components with passive optical elements and detector arrays could extend the low-power platform to dense industrial, infrastructure and robotics sensing.
Can this light-speed optical sensor overcome the high costs and manufacturing hurdles needed for mass market adoption?
With sensors now 'thinking' at light speed, is the era of electronic edge computing for robotics and IoT already over?
AOFS-IC: Ultrafast All-Optical Fiber Sensing with <3 ns Demodulation and Sub-Nano Strain Resolution
Overview
The AOFS-IC system marks a major leap in fiber sensing by overcoming the long-standing limitations of traditional optical fiber sensors, which suffer from latency and high power use due to electronic signal processing. By moving all sensing and demodulation into the optical domain, AOFS-IC uses a scattering medium and a diffractive optical network to map physical changes directly to light intensity. This allows for immediate, ultrafast readout without electronics, enabling highly efficient and accurate sensing. The result is a new era of fiber sensing that is faster, simpler, and more energy-efficient than ever before.