Original title: Guided Lamb-wave monitoring for structural ice detection: Gaussian-process-based detection under environmental variability
Nicola Roveri, Lorenzo Stagi, Simone Tedeschi, Antonio Concilio +1 · August 2026
Ice building up on aircraft and wind turbines can create severe safety hazards, but typical de-icing systems use massive amounts of energy. To solve this, researchers developed a low-power detection method that uses piezoelectric transducersSmall devices that can both send out a sound wave when electricity is applied and detect sound waves as electrical signals., which are devices that turn electrical signals into physical vibrations and vice versa. Their approach sends guided Lamb wavesSound waves that travel through thin materials like metal plates, changing speed and shape if something like ice forms on the surface., a type of acoustic wave traveling through thin materials, directly across a surface. Ice changes these wave patterns. By tracking wave changes over time using normalized autocorrelation and logarithmic envelopes, the system measures distortions caused by ice. To prevent temperature changes from triggering false alarms, they used a statistical tool called Gaussian Process Regression to learn normal baseline behavior. When tested on an aluminum plate, this method reached 100% detection accuracy in three sensor setups and 96% in a fourth. It reliably spotted ice while handling temperature swings from +18°C down to -18 °C.
Imagine you tap a wooden table every morning and listen to the sound it makes. On a warm day it sounds one way, on a cold day slightly different, but you learn the range of normal sounds. Now suppose one morning it sounds like the table is muffled—that could be ice on the surface, which changes the noise. But because you have learned what to expect in different temperatures, you can tell if the change is just from the cold or actually from ice. In this paper, the system taps the structure with sound waves and uses a smart model that has learned what normal echoes look like across temperatures, so it can spot when ice is causing an abnormal echo.
Ice buildup on aircraft wings and wind turbine blades is dangerous, but current de-icing systems use a lot of energy. This research shows a way to detect ice using low-power sensors that send sound waves through the structure, so you only turn on energy-heavy de-icing when it's actually needed. That saves fuel on planes and keeps wind turbines running safely in cold weather.
The study was validated only on a simple aluminum plate, not on actual aircraft wings or wind turbine blades, which have more complex shapes and materials. The experiments covered a temperature range from +18°C to -18°C, but real conditions may involve wider swings or rapid changes. The abstract also notes a trade-off between sensitivity and thermal robustness as the signal window length changes, suggesting the method may need tuning for different ice detection needs.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Guided Lamb-wave monitoring for structural ice detection: Gaussian-process-based detection under environmental variability — Nicola Roveri, Lorenzo Stagi, Simone Tedeschi et al. — Ultrasonics — doi:10.1016/j.ultras.2026.108271
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper.