Digital twin of fiber optics: Perm scientists learn to predict internet and navigation failures
Imagine: an aircraft is on its landing approach, and suddenly the navigation system starts making errors. Or in the middle of a live broadcast, a data center loses data packets, and the transmission is interrupted. The culprit could be a tiny optical fiber — that same glass thread that carries light and information around the world. Or rather, not the fiber itself, but its protective plastic coating, which behaves unpredictably during sharp temperature changes.
The problem is as old as the hills: polymers soften in heat and harden in cold. For an ordinary internet cable, this is not critical, but when it comes to high-precision sensors in aviation, space, or medicine, any temperature fluctuation can distort the signal. Engineers have to play it safe — making protective layers thicker and heavier, which increases the cost of the structure and adds extra weight. For a space satellite, every gram counts, and for an aircraft, it means extra fuel consumption. But the risk of failure still remains, because no one knew how to accurately predict the coating’s behavior under real-world conditions.
Until today. Scientists at Perm Polytechnic have created the world’s first digital twin of optical fiber that predicts its behavior in extreme temperatures — from minus 110 to plus 120 degrees Celsius — with up to 90% accuracy. And most importantly, it considers not just the temperatures themselves, but the rate of change. Because it is precisely a sharp fluctuation, for example, from a warm hangar to the cold outside an aircraft, that creates maximum stress on the material.
To create such a model, scientists spent several years painstakingly experimenting: stretching samples at different frequencies, heating and cooling them with liquid nitrogen, recording thousands of parameters. How the material deforms, how it accumulates elastic energy, how it dissipates heat. The result was a mathematical model of a two-layer polymer coating — a digital twin that behaves exactly like the real material.
The development was tested on a real-world task — fiber optics for gyroscopes that determine the orientation of drones, aircraft, and satellites. They simulated a sharp temperature drop from plus 60 to minus 60 degrees Celsius — precisely the kind of contrast that compresses and stretches the fiber from within most severely. The program calculated how the glass and polymers press against each other, where stresses arise, and how the signal changes. The result was a real “operational map,” looking at which an engineer can immediately see weak points and, at the design stage, adjust the shape of the protective layers or fine-tune signal processing.
The result is impressive: the accuracy of fiber optic sensors in extreme conditions can be increased by 25–40%, and the devices themselves can be made lighter and cheaper by removing excess protection where it is not needed. For aviation and space, this is not just about economy — it is a matter of safety and competitiveness.
