Aircraft have traditionally been designed with a separation between structure, sensing and control. Sensors measure parameters such as pressure, temperature, acceleration and strain, while flight-control systems interpret this information and command aerodynamic surfaces.

A new research direction is attempting to integrate these functions directly into the aircraft surface.

A 2026 review published in Research examined the development of AI-enabled flexible sensing skins for aircraft. The researchers describe flexible sensing skins as lightweight, conformal systems capable of covering large surface areas and providing distributed information about an aircraft's surrounding environment and structural state.

The idea is sometimes associated with the concept of “fly-by-feel.”

Traditional aircraft rely heavily on discrete sensors. A sensing skin could instead create a distributed sensory layer across the aircraft.

Imagine an aircraft wing covered with thousands of miniature sensing elements. Instead of measuring aerodynamic pressure at only a few locations, the aircraft could obtain spatially distributed information about pressure, strain, temperature or flow-related phenomena.

Artificial intelligence could then interpret this high-dimensional dataset.

This is particularly interesting for next-generation aircraft because aerodynamic behaviour is dynamic. During manoeuvres, turbulence or changing flight conditions, the distribution of pressure and structural strain can change rapidly.

An intelligent surface could potentially identify patterns associated with changing aerodynamic conditions and communicate them to flight-control systems.

The concept also has implications for structural health monitoring. A distributed sensing layer could potentially detect changes associated with damage, deformation or unusual loading.

However, creating such a system presents major engineering challenges. Flexible sensors must remain lightweight and mechanically robust. They must tolerate temperature variation, aerodynamic loading, vibration, moisture and long-term environmental exposure.

Data processing is another major problem.

A large aircraft surface could generate enormous quantities of sensor information. Conventional control systems may not be designed to process such high-dimensional datasets in real time. Machine-learning architectures may therefore become increasingly important.

The research direction represents a convergence of aerodynamics, materials science, flexible electronics, artificial intelligence, structural mechanics and aerospace control.

The most interesting aspect is that the aircraft surface could become more than a passive aerodynamic boundary.

It could become a computational interface between the aircraft and its environment.

Rather than asking only whether an aircraft can fly efficiently, future aerospace engineers may increasingly ask whether an aircraft can sense, interpret and respond to its aerodynamic environment through its own structure.

This would represent a transition from conventional aircraft instrumentation toward increasingly integrated, intelligent aerospace structures.

Reference:

• Ji et al., Research (2026), “AI-Enabled Flexible Sensing Skin for Next-Generation Aircraft: Toward Embodied Intelligence.”