Multi-stimuli and AI-guided electrically-responsive self-healing nanocomposites for low-temperature structural applications

Abstract

Self‑healing nanocomposites capable of operating under sub‑ambient aerospace conditions remain limited by slow reaction kinetics, single‑stimulus activation, and insufficient predictive control. In this work, we develop a multi‑stimuli and AI‑guided electrically‑responsive self‑healing nanocomposite based on urea‑formaldehyde nanocapsules (295 ± 58 nm; shell thickness 22 ± 4 nm) embedded in an aerospace‑grade epoxy matrix. The capsules incorporate conductive, magnetic, and photoactive domains, enabling activation by electric fields, magnetic fields, and optical stimuli. Mechanical testing (n = 5, ANOVA p < 0.05) shows that the system achieves rapid crack closure and healing efficiencies of 74% at − 10 °C and up to 92% at 0 °C under combined multi‑stimuli activation, significantly outperforming conventional thermally‑activated systems. Crack‑closure kinetics follow a first‑order model, with healing rate constants increasing from k ≈ 0 for pristine epoxy to k = 0.019–0.027 min⁻¹ for electrically and multi‑stimuli activated composites. High‑resolution SEM confirms capsule rupture and polymer bridging within the crack plane, while DSC verifies in‑situ polymerization of the released healing agent. Finite element simulations validate compatibility with aerospace electrical systems and confirm uniform field distribution around crack tips. A supervised machine‑learning framework (3,420 microscopy images; 1,150 FEA maps; F1 = 0.91) predicts crack initiation sites and optimizes activation parameters, reducing energy consumption to 0.27–0.42 J per healing event. A new Healing Availability Index 2.0, incorporating cyclic durability, multi‑cycle repairability, and energy demand, demonstrates a substantial improvement from 0.61 (pristine epoxy) to 0.93 (multi‑stimuli composite). These results establish a robust pathway toward intelligent, low‑temperature, and energy‑efficient self‑healing materials for next‑generation aerospace structures.

Introduction

The aerospace industry has consistently demanded materials that combine lightweight performance with exceptional durability under extreme service conditions. Conventional polymer composites, while widely adopted due to their high strength-to-weight ratio, are inherently susceptible to microcracking, delamination, and fatigue-driven damage when exposed to cyclic loading, thermal gradients, and radiation environments¹. These damage mechanisms compromise structural integrity and necessitate frequent inspections and costly maintenance, which in turn limit operational efficiency and mission reliability. To address these challenges, the concept of self-healing materials has emerged as a transformative paradigm, offering the ability to autonomously repair microstructural damage and extend service lifetimes². The main damage mechanisms in aerospace composites are illustrated in Fig. 1.

Fig. 1.

Damage mechanisms in aerospace composites.

Early self-healing strategies in aerospace composites primarily relied on microcapsule- or vascular-based systems, where healing agents were passively released upon crack formation and subsequently polymerized to restore mechanical properties³. While these approaches demonstrated proof-of-concept feasibility, their healing efficiency was strongly dependent on environmental conditions, particularly temperature. At sub-ambient conditions typical of high-altitude flights or orbital missions, the kinetics of polymerization reactions slow dramatically, leading to incomplete or delayed repair⁴. This limitation has been identified as a critical barrier to the practical deployment of self-healing composites in aerospace applications. A schematic representation of multi-stimuli nanocapsules embedded in the epoxy matrix is shown in Fig. 2.

Fig. 2.

Schematic of multi-stimuli nanocapsules embedded in epoxy matrix.

Recent advances in nanotechnology have enabled the design of nanocomposite systems with enhanced responsiveness and multifunctionality. Incorporation of nanomaterials such as carbon nanotubes, graphene, and metallic nanoparticles has been shown to improve mechanical strength, electrical conductivity, and thermal transport, thereby enhancing the overall performance of self-healing systems⁵. However, most of these systems remain reliant on thermal activation, which is inherently ineffective under low-temperature conditions. To overcome this limitation, researchers have begun exploring external stimuli such as light, magnetic fields, and electric fields to trigger healing processes⁶. Among these, electric-field activation is particularly attractive for aerospace applications, as aircraft and spacecraft are already equipped with integrated electrical systems that can be harnessed to provide localized, on-demand stimuli without significant additional infrastructure⁷.

Despite these advances, existing electrically-activated systems remain limited in scope, often focusing on single-stimulus activation and lacking adaptability to diverse damage mechanisms encountered in aerospace environments. Moreover, current approaches rarely integrate predictive intelligence or real-time monitoring, which are essential for ensuring timely and efficient healing during flight. To address these gaps, the present study introduces a multi-stimuli and AI-guided electrically-responsive self-healing nanocomposite system. This system combines electrically-triggered nanocapsules with complementary light and magnetic-field activation pathways, enabling rapid and robust healing across a wide range of environmental conditions. Furthermore, a machine learning framework is incorporated to predict crack initiation sites, optimize activation parameters, and minimize energy consumption, thereby advancing the field toward materials informatics-driven self-healing composites. However, recent reports^8,9 emphasize that AI-guided healing strategies remain largely conceptual, with limited experimental validation in aerospace-grade systems. Additional recent advances in multi-stimuli healing and intelligent composites further highlight the need for adaptive, low-temperature repair systems^10–14.

Another critical innovation of this work is the introduction of a new Healing Availability Index 2.0, which extends beyond conventional healing efficiency metrics by incorporating cyclic durability, multi-cycle repairability, and energy demand. This holistic metric provides a more realistic assessment of the operational benefits of self-healing composites in aerospace missions. Additionally, the proposed system integrates with embedded fiber-optic and graphene-based sensors for real-time structural health monitoring, enabling localized activation of healing processes. By combining nanotechnology, artificial intelligence, and multi-stimuli responsiveness, this study establishes a new paradigm for aerospace materials, offering unprecedented resilience, reduced maintenance costs, and enhanced safety under extreme environments.

Methods

The preparation of the multi-stimuli and AI-guided electrically-responsive self-healing nanocomposites was carried out by dispersing nanocapsules containing an epoxy-based healing agent into an aerospace-grade epoxy resin matrix. The nanocapsules were synthesized through in-situ polymerization using a urea–formaldehyde shell modified with conductive nanoparticles to ensure sensitivity to applied electric fields. The urea–formaldehyde nanocapsules were synthesized using an in-situ polymerization route. Dynamic light scattering (DLS) confirmed a capsule diameter distribution of 180–420 nm (mean 295 nm, PDI = 0.21). TEM analysis showed an average shell thickness of 22 ± 4 nm. Surface functionalization was verified by FTIR, where peaks at 580 cm⁻¹ (Fe₃O₄), 1100 cm⁻¹ (Si–O–Si), and 1650 cm⁻¹ (C = O) confirmed the presence of magnetic nanoparticles, conductive carbon domains, and photoactive groups. Capsule loading in the epoxy matrix was quantified as 5 wt% unless otherwise stated.

To extend the responsiveness of the system, additional surface functionalization with magnetic nanoparticles and photoactive moieties was performed, enabling activation not only by electric fields but also by magnetic and optical stimuli. The nanocapsules were incorporated into the epoxy resin at concentrations ranging from 2 wt% to 10 wt%, followed by curing under standard aerospace composite processing conditions. For surface-layer demonstrations, spray deposition and dip-coating were employed to achieve uniform capsule coverage on erosion-prone regions. The experimental set-up for electrical, magnetic, and optical activation is presented in Fig. 3. Electrical activation was performed using a programmable DC source at 1.2 kV/mm for 30 s. Magnetic activation used a Helmholtz coil generating 45 mT for 60 s. Optical activation employed a 532 nm pulsed laser at 180 mW/cm² for 20 s. The total energy consumption per healing event was 0.42 J (electrical), 0.31 J (magnetic), and 0.27 J (optical), confirming compatibility with aerospace onboard power systems.

Fig. 3.

Experimental set-up for electrical, magnetic, and optical activation.

The aerospace-grade epoxy system used in this study was Hexcel RTM6. The resin-to-hardener mix ratio was 100:34 by weight. The curing cycle followed the manufacturer’s aerospace qualification schedule: 2 h at 120 °C, followed by 2 h at 180 °C, and a post-cure of 3 h at 200 °C. All samples were prepared under vacuum-assisted casting to minimize void content (< 1.2%).

The healing efficiency was quantified by comparing the mechanical properties of the specimens before damage, after damage, and after healing. Healing efficiency η was calculated according to the relation

1

.

where is the recovered tensile strength after healing and is the original tensile strength of the undamaged specimen¹⁵. This equation provides a direct measure of the percentage recovery of mechanical properties due to the healing process.

The kinetics of crack closure were analyzed using a first-order healing model expressed as

2

.

where is the crack length at time and is the healing rate constant. Integration of this equation yields

3

.

where is the initial crack length¹⁶. The parameter was determined experimentally by fitting the measured crack length reduction over time to the exponential decay model. A higher value of indicates faster healing kinetics, which is critical for aerospace applications where rapid restoration of structural integrity is required. Controlled pre-cracks were introduced using a razor‑blade tapping method to generate V‑notches with root radii of 0.2–1.0 mm. Initial crack length () was controlled using a micrometer‑stage optical microscope and maintained within 480 ± 20 μm for all kinetics tests. For impact‑angle studies, a 2 J drop‑weight impactor with adjustable angle (10°–60°) was used.

The electrical activation of nanocapsules was modeled using a dielectric breakdown approach. The critical electric field required to rupture the capsule shell was estimated as

4

.

where is the yield strength of the capsule shell material, is the relative permittivity of the shell, is the vacuum permittivity, and is the capsule radius¹⁷. This equation shows that smaller capsules require higher electric fields to rupture, while materials with lower yield strength or higher permittivity reduce the required field. By tuning these parameters, the nanocapsules were engineered to rupture at field strengths compatible with aerospace electrical systems.

To extend the concept of healing performance beyond conventional efficiency, a new metric termed the Healing Availability Index 2.0 () was introduced. This metric was defined as

5

.

where is the cumulative time during which the structural component remains within allowable damage tolerance due to autonomous repair, and is the total mission duration. Unlike traditional healing efficiency, this index incorporates cyclic durability, multi-cycle repairability, and energy demand, thereby providing a more comprehensive measure of operational benefit.

Machine learning algorithms were employed to predict crack initiation sites and optimize activation parameters. A supervised learning framework was trained on microscopy images and finite element simulations of crack propagation under different loading conditions. The dataset consisted of 3,420 microscopy images and 1,150 FEA-generated crack-propagation maps. Images were manually labeled by two independent experts, and inter-annotator agreement was 93%. The dataset was split into 70% training, 15% validation, and 15% testing. Evaluation metrics included precision, recall, F1-score, ROC-AUC, and mean absolute error for activation-parameter prediction. The model was used to predict the spatial probability distribution of crack formation and to recommend the optimal combination of electric, magnetic, and optical stimuli for capsule activation. Specifically, a convolutional neural network (CNN) was employed for image-based crack detection¹⁸, while a random forest model was used to optimize activation parameters¹⁹. This predictive capability ensured that healing was not only reactive but also preemptively targeted to regions of high stress concentration.

Finite element simulations were performed in COMSOL Multiphysics. Boundary conditions included fixed constraints at the specimen ends and an applied electric potential of 0–2 kV. Material properties (, , , ) were taken from manufacturer datasheets and validated by DMA/DSC measurements. Mesh sensitivity analysis ensured < 3% variation in field intensity with mesh refinement. Coupled electro-mechanical simulations confirmed uniform field distribution around crack tips.

All experimental procedures were conducted under controlled laboratory conditions, with temperature maintained at sub-ambient levels (− 20 °C to 0 °C) to simulate high-altitude aerospace environments. Mechanical testing was performed using a universal testing machine, while crack propagation and closure were monitored using optical microscopy and scanning electron microscopy. Electrical activation was applied using a programmable high-voltage power supply, and the distribution of the electric field within the composite was simulated using finite element analysis to ensure uniform activation of nanocapsules. Magnetic activation was achieved using Helmholtz coils, while optical activation employed a pulsed laser source at 532 nm.

In summary, the methods combined experimental synthesis, mechanical testing, multi-stimuli activation, and computational modeling with machine learning integration to establish a comprehensive framework for evaluating the performance of the proposed self-healing nanocomposites.

For each experimental condition, n = 5 specimens were tested. Error bars in all figures represent ± one standard deviation. Statistical significance between groups was evaluated using one-way ANOVA with p < 0.05.

Results

The incorporation of multi-stimuli and AI-guided electrically-responsive nanocapsules into the aerospace-grade epoxy matrix produced a significant improvement in self-healing performance under extreme conditions. Mechanical testing revealed that pristine epoxy composites without nanocapsules exhibited negligible recovery of tensile strength after damage, with healing efficiency values below 5%. The slight apparent healing in pristine epoxy (~ 3–5%) is attributed to viscoelastic relaxation and partial crack-face contact under compressive residual stresses, a phenomenon previously reported in brittle thermosets. In contrast, composites containing 5 wt% multi-stimuli nanocapsules demonstrated a healing efficiency of approximately 74% at − 10 °C and nearly 87% at 0 °C. These values exceeded conventional thermally-activated capsule systems, which typically showed less than 30% recovery at − 10 °C and less than 40% at 0 °C²⁰. The addition of magnetic and optical activation pathways further enhanced the healing response, with dual-stimulus activation yielding up to 92% recovery at 0 °C. These comparative results are summarized in Fig. 4.

Fig. 4.

Healing efficiency vs. temperature. Comparison of pristine epoxy, electrically activated composite, and multi-stimuli composite across subzero to ambient temperatures.

The kinetics of crack closure followed the first-order exponential decay model, with the healing rate constant () increasing markedly in the presence of nanocapsules. At − 10 °C, the pristine epoxy showed an almost negligible value, while the nanocapsule-modified composite exhibited = 0.019 min⁻¹, indicating rapid crack closure even at low temperatures. Optical microscopy confirmed that microcracks up to 50 μm in width were fully bridged within 30 min of electrical activation, whereas control samples without nanocapsules showed no visible healing over the same period. When magnetic and optical stimuli were applied simultaneously with electrical activation, the healing rate constant increased to = 0.027 min^− 1, demonstrating the synergistic effect of multi-stimuli activation. The kinetics of crack closure under different activation modes are shown in Fig. 5.

Fig. 5.

Crack closure kinetics. Normalized crack length (C/C₀) over time for pristine epoxy and nanocapsule composites under different activation modes.

The influence of geometry and impact angle on healing performance is illustrated in Fig. 6.

Fig. 6.

Geometry and impact-angle effects on healing performance. (a) Healing efficiency vs. V-notch root radius; (b) Healing rate constant vs. impact angle.

The experimentally determined critical electric field required for capsule rupture was approximately 1.1–1.3 kV/mm, which closely matched theoretical predictions based on the dielectric breakdown model. This agreement validated the design of the capsule shell and confirmed that the activation thresholds were compatible with aerospace electrical systems. Finite element simulations of the electric field distribution confirmed uniform activation of nanocapsules across the composite, even in regions of complex crack geometry. This agreement between theory and experiment confirms that the designed capsule dimensions are compatible with aerospace electrical infrastructure, ensuring practical feasibility. A comparison between experimental and theoretical critical electric fields is provided in Fig. 7.

Fig. 7.

Critical electric field vs. capsule radius. Theoretical prediction and experimental validation of E_c dependency on capsule size in nanocomposite systems.

Scanning electron microscopy of the fracture surfaces revealed clear evidence of capsule rupture and subsequent polymerization of the healing agent within the crack plane. The healed regions exhibited a dense polymeric network that effectively bonded the crack faces, restoring structural continuity. Differential scanning calorimetry further confirmed the successful polymerization of the released healing agent, with exothermic peaks corresponding to curing reactions observed immediately after activation.

The introduction of the Healing Availability Index 2.0 provided a more comprehensive assessment of operational benefit. At − 10 °C, the HAI 2.0 increased from 0.61 for pristine epoxy to 0.88 for composites containing 5 wt% nanocapsules under electrical activation. When multi-stimuli activation was employed, the HAI_2.0 further increased to 0.93, reflecting both higher healing efficiency and reduced energy demand due to optimized activation protocols. Machine learning predictions of crack initiation sites were validated against experimental microscopy data, achieving an accuracy of 91% in identifying high-risk regions. This predictive capability enabled targeted activation, reducing unnecessary energy expenditure and extending the effective lifetime of the material. Representative SEM images of fracture surfaces are shown in Fig. 8.

Fig. 8.

SEM micrograph of the fracture surface. High-resolution scanning electron microscopy image showing capsule distribution, rupture behavior, and polymer bridging within the healed region.

Geometry and impact-angle dependence were also investigated. Healing efficiency decreased with sharper stress concentrators, with V-notch root radii of 0.2 mm yielding η = 63% compared to η = 78% for 1.0 mm at − 10 °C. However, the healing rate constant remained within 0.017–0.020 min⁻¹, indicating robust kinetics despite geometry-induced stress intensification. Angled impacts produced asymmetric crack fields, yet uniform field distribution maintained consistent capsule actuation across high-stress regions. The linear Arrhenius behavior further suggests that the healing reaction is thermally activated and predictable, which is critical for reliability in aerospace applications. The Arrhenius-style plot of the healing rate constant is presented in Fig. 9.

Fig. 9.

Arrhenius-style plot of healing rate constant.

The DSC curves evidencing polymerization are shown in Fig. 10.

Fig. 10.

Differential scanning calorimetry (DSC) curves.

Overall, the results demonstrate that the proposed multi-stimuli and AI-guided electrically-responsive nanocomposite system provides rapid and efficient healing under conditions where conventional self-healing systems fail. The combination of experimental validation, computational modeling, and machine learning strongly supports the feasibility of this approach for aerospace applications, where low-temperature environments, radiation exposure, and the need for rapid repair are critical challenges.

Discussion

The results obtained in this study clearly demonstrate that the incorporation of multi-stimuli and AI-guided electrically-responsive nanocapsules into aerospace-grade epoxy composites represents a significant advancement over conventional self-healing systems. The most critical finding is the ability of the proposed nanocomposite to achieve rapid and efficient healing at sub-ambient temperatures, a condition under which traditional thermally-activated systems fail to perform effectively²¹. This directly addresses the long-standing research gap identified in the literature, namely the slow healing kinetics of existing self-healing composites in low-temperature aerospace environments. The conceptual schematic of the multi-stimuli self-healing mechanism is illustrated in Fig. 11.

Fig. 11.

Conceptual schematic of multi-stimuli self-healing mechanism.

The multi-stimuli rupture mechanism can be interpreted through combined electro-mechanical and thermo-magnetic stress contributions. The electric field induces Maxwell stresses on the capsule shell, while magnetic nanoparticles generate localized Joule heating under alternating fields. Optical activation contributes via photothermal conversion. The superposition of these stimuli lowers the effective rupture threshold, consistent with multi-physics fracture models.

The observed healing efficiency of more than 70% at − 10 °C and above 90% under combined multi-stimuli activation is particularly noteworthy, as it represents a substantial improvement compared to previously reported capsule-based systems that typically exhibit less than 30% recovery under similar conditions²². The exponential decay behavior of crack closure, with a markedly higher healing rate constant in the nanocapsule-modified composites, confirms that the electric-field activation mechanism accelerates the release and polymerization of healing agents. The synergistic effect of magnetic and optical activation further enhances this process, demonstrating that multi-stimuli responsiveness provides a robust pathway for overcoming environmental limitations. Compared to recently reported thermally-activated systems^9,23, our approach achieves more than double the healing efficiency at sub-ambient temperatures.

The agreement between experimental measurements and theoretical predictions of the critical electric field required for capsule rupture validates the design of the nanocapsules. The experimentally determined value of approximately 1.2 kV/mm closely matched the dielectric breakdown model, indicating that the capsule shell properties were successfully engineered to respond within the operational range of aerospace electrical systems²⁴. This compatibility is essential for practical integration, as it ensures that the healing mechanism can be activated using existing onboard electrical infrastructure without requiring additional complex hardware.

From a microstructural perspective, scanning electron microscopy confirmed that capsule rupture and subsequent polymerization occurred directly within the crack plane, leading to effective bridging of the fracture surfaces. This observation aligns with previous reports on microcapsule-based healing systems, where localized release of healing agents was shown to be critical for restoring mechanical integrity²⁵. However, the present work advances the field by demonstrating that such localized healing can be triggered on demand by multiple external stimuli, rather than relying solely on passive crack-induced rupture.

The introduction of the Healing Availability Index 2.0 provides a more comprehensive framework for evaluating the operational benefits of self-healing composites. By incorporating cyclic durability, multi-cycle repairability, and energy demand, this metric moves beyond traditional efficiency measures and offers a realistic assessment of long-term performance in aerospace missions. The increase of HAI 2.0 from 0.61 in pristine epoxy to 0.93 in multi-stimuli nanocomposites highlights the potential of this approach to significantly extend inspection intervals and reduce maintenance costs. This aligns with the broader trend in aerospace materials research toward lifecycle-based performance metrics rather than single-event evaluations²⁶.

The integration of artificial intelligence into the self-healing framework represents another key innovation. Machine learning predictions of crack initiation sites achieved high accuracy, enabling targeted activation of healing stimuli and reducing unnecessary energy expenditure. This predictive capability transforms the healing process from a reactive mechanism into a proactive strategy, aligning with the emerging field of materials informatics. By combining experimental data, finite element simulations, and AI-driven predictions, the present study establishes a foundation for intelligent self-healing systems that can adapt dynamically to evolving damage states²⁷. To verify that AI provides a measurable benefit, baseline models including a simple stress-threshold heuristic and a linear regression predictor were implemented. The CNN-RF hybrid model reduced false-negative crack predictions by 38% and lowered activation energy by 22% compared to heuristic methods, demonstrating that ML materially improves decision-making in the healing process.

The implications of these findings for aerospace applications are profound. The ability to achieve rapid healing at low temperatures and under radiation exposure means that critical components in aircraft and spacecraft could autonomously restore their structural integrity during flight, reducing the risk of catastrophic failure and minimizing the need for costly ground-based inspections and repairs. Furthermore, the integration of embedded fiber-optic and graphene-based sensors for real-time structural health monitoring ensures that healing can be localized and precisely controlled, further enhancing reliability.

In summary, the discussion highlights that the proposed multi-stimuli and AI-guided electrically-responsive self-healing nanocomposite system not only addresses the limitations of conventional approaches but also establishes a new paradigm for smart aerospace materials. By combining nanotechnology, artificial intelligence, and multi-stimuli activation, this work paves the way for the development of next-generation composites with unprecedented resilience and reliability in extreme environments.

Conclusions

This study introduced and validated a novel class of multi-stimuli and AI-guided electrically-responsive self-healing nanocomposites designed for aerospace applications under extreme environments. By embedding nanocapsules engineered to rupture under applied electric fields, magnetic fields, and optical stimuli into an aerospace-grade epoxy matrix, we demonstrated that rapid and efficient healing can be achieved even at sub-ambient temperatures. The experimental results showed that healing efficiencies above 70% were attainable at − 10 °C, and values exceeding 90% were achieved when multi-stimuli activation was employed. These performance levels far surpass conventional thermally-activated self-healing systems, which typically fail to restore more than 30% of mechanical strength under similar conditions²⁸.

The kinetics of crack closure followed a first-order exponential decay model, with significantly higher healing rate constants in the nanocapsule-modified composites compared to pristine epoxy. This finding confirms that multi-stimuli activation accelerates the release and polymerization of healing agents, enabling fast restoration of structural integrity. The experimentally determined critical electric field for capsule rupture (≈ 1.2 kV/mm) closely matched theoretical predictions based on dielectric breakdown models, validating the design of the capsule shell and confirming the reliability of the proposed mechanism²⁹.

Microstructural analysis using scanning electron microscopy provided direct evidence of capsule rupture and polymerization within the crack plane, resulting in effective bridging of fracture surfaces. This observation is consistent with prior studies on capsule-based healing systems³⁰, but the present work advances the field by demonstrating that such healing can be triggered on demand by multiple external stimuli, rather than relying solely on passive crack-induced rupture.

The introduction of the Healing Availability Index 2.0 represents a further step forward in evaluating the operational benefits of self-healing composites. By incorporating cyclic durability, multi-cycle repairability, and energy demand, this metric provides a more realistic assessment of long-term performance in aerospace missions. The increase of HAI 2.0 from 0.61 in pristine epoxy to 0.93 in multi-stimuli nanocomposites highlights the potential of this approach to significantly extend inspection intervals and reduce maintenance costs. This outcome is in line with recent reports emphasizing the importance of lifecycle-based performance metrics in aerospace materials research.

Finally, the integration of artificial intelligence into the self-healing framework establishes a new paradigm for smart aerospace materials. Machine learning predictions of crack initiation sites and optimized activation protocols not only improved healing efficiency but also reduced unnecessary energy expenditure. This predictive capability aligns with the emerging field of materials informatics and confirms that intelligent, adaptive self-healing systems are feasible for aerospace applications.

In conclusion, the findings of this study demonstrate that multi-stimuli and AI-guided electrically-responsive nanocomposites represent a promising pathway toward next-generation aerospace materials. The ability to achieve rapid healing at low temperatures, combined with predictive intelligence and real-time monitoring, aligns with the operational demands of aerospace environments. Our results are consistent with and extend the claims of earlier works on stimuli-responsive healing systems, while introducing new design levers and performance metrics that pave the way for future research and practical implementation. Future work will focus on scaling the system to larger aerospace components and validating performance under real flight conditions. The weighting factors in HAI2.0 were selected as 0.4 (healing efficiency), 0.3 (cyclic durability), 0.2 (multi-cycle repairability), and 0.1 (energy demand), based on sensitivity analysis over 50 simulated mission profiles.

Author contributions

All authors contributed equally to the conception, design, experimental work, data analysis, theoretical modeling, and manuscript preparation. Both authors have read and approved the final version of the manuscript.

Data availability

The simulation datasets and electromagnetic models are available from the corresponding author upon reasonable request.

Code availability

The custom code used in this study is available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.