The patent landscape: who is leading embedded blade sensing

Vestas Wind Systems A/S, Siemens Gamesa Renewable Energy A/S, and General Electric Company hold the dominant positions in embedded sensing for wind turbine blade load monitoring, based on a dataset of approximately 60 patent records spanning 2010 to 2025 across jurisdictions including EP, US, WO, GB, IN, and AU. The breadth of this portfolio reflects the commercial urgency of the problem: blade structural failures represent some of the most costly and dangerous failure modes in modern wind energy systems, and the industry has responded with sustained, multi-decade investment in in-situ monitoring technology.

Vestas is the most prolific patent holder in blade-specific load sensing. Their portfolio spans internally mounted root insert strain sensors (GB, 2010; EP, 2019), axial force correction algorithms (EP, 2022), temperature-compensated load measurement (EP, 2021), fault detection in load sensors (EP, 2023), optical block sensor units (EP, 2024), and composite shell insert inspection methods (WO, 2026). The consistent technical thread across this portfolio is the correction of raw sensor measurements for confounding variables — temperature, axial force, and blade dynamics — to produce accurate real-time load parameters suitable for individual pitch control (IPC) feedback.

Siemens Gamesa’s portfolio emphasises multi-functional monitoring architectures and composite-specific sensing, including a dual-use lightning protection and structural health monitoring network (EP, 2022 and US, 2022), delamination detection via conductive fasteners (EP, 2021), embedded optical fibers for lifetime load monitoring (EP, 2022), and electrical time-domain reflectometry (TDR)-based strain control (EP, 2025). General Electric’s patents address broader structural health monitoring (SHM) system architecture — sensor-to-controller monitoring pipelines (US, 2011), optical fiber stress monitoring for blades (EP, 2017), and additive manufacturing of tower structures with embedded reinforcement sensing elements (IN, 2020).

Academic contributions to this field originate from institutions including the Technical University of Denmark (DTU/Risø), ETH Zürich, Xi’an Jiaotong University, IFSTTAR/Université Gustave Eiffel, and the National Institute of R&D for Technical Physics (Romania). DTU/Risø appears most frequently in the academic literature, covering damage mechanism-based SHM, distributed accelerometer-based vibration monitoring, sensor fault detection in IPC systems, and vibration benchmark studies. According to WIPO, wind energy-related patent filings have accelerated significantly over the past decade, reflecting the sector’s transition toward larger, more structurally complex turbines that demand more sophisticated monitoring.

Sensor modalities and how they integrate with composite laminates

Fiber optic sensors — particularly Fiber Bragg Gratings (FBGs) and distributed Brillouin-based systems — are the most extensively patented and studied technology for in-situ load monitoring in wind turbine blade composites. The core principle involves embedding optical fibers either as additions to, or substitutes for, structural reinforcing fibers within the glass- or carbon-fiber-reinforced plastic (GFRP/CFRP) laminate. Siemens Gamesa’s 2022 patent demonstrates that optical fibres embedded in the blade measure loads on and within the blade surface throughout its complete lifetime from manufacturing to decommissioning. General Electric’s EP 2017 patent extends this: optical glass fibers embedded on or within the blade surface continuously transmit light, and a loss of transmission directly correlates to exceedance of the fiber’s tensile strength — providing an unambiguous structural load state signal.

The physical integration of FBGs into composite plies is not trivial. Research from IFSTTAR (Université Gustave Eiffel) on fatigue behavior of smart composites with distributed fiber optic sensors for offshore applications (2021) establishes that placement geometry — linear versus sinusoidal routing within plies — significantly affects coverage of critical stressed zones such as trailing edges and spar webs. An earlier IFSTTAR study (2019) addresses the formation of resin-rich “penny-shape” defects around embedded fiber optic sensor cross-sections and advocates dual-sinusoidal placement to maximize inter-ply and bonding-zone coverage while minimizing sensor-induced material defects. Research from Nature-indexed journals has further confirmed that distributed sensing architectures — including Brillouin-based systems — enable shape reconstruction and damage detection in large-scale composite structures, with direct applicability to multi-meter wind turbine blades.

Piezoceramic, acoustic emission, and impedance-based sensing

Piezoceramic transducers serve dual roles in smart composite wind turbine blades — as actuators that excite guided acoustic waves into the structure and as sensors that receive and analyse those waves for damage-related signal changes. Research from Harbin Institute of Technology (2013) developed a complete wireless sensor network (WSN) architecture where one wireless node drives an embedded piezoceramic patch to generate guided waves, while remaining nodes detect wave responses across distributed blade locations. Damage indices derived from wavelet packet analysis of these signals were validated through both static loading tests and wind tunnel experiments.

Siemens Gamesa’s 2022 study on structural health monitoring of a 52-meter wind turbine blade deployed strain gauges, acoustic emission sensors, distributed accelerometers, and an active vibration monitoring system during fatigue testing, tracking the propagation of artificially introduced laminate cracks into delaminations. This multi-modal approach illustrated both the complementary nature and the individual limitations of each sensing modality. A comprehensive review from Ho Chi Minh City University of Technology (2022) on piezoelectric impedance-based SHM highlights that impedance-based methods are robust to early-stage failures, cost-effective, and amenable to real-time damage assessment — though the review notes that very few studies had yet applied this technique specifically to wind turbine structures at the time of publication.

Strain gauges, wireless sensor networks, and FEM-guided placement

Resistive strain gauges remain a practical baseline sensor for blade root and distributed blade load measurement. Vestas Wind Systems’ GB 2010 patent describes a strain gauge embedded within a root insert, fixedly connected so that loads are transmitted directly between the insert and the sensing element, enabling precise root bending moment measurement. The National Institute of R&D for Technical Physics (Romania) has contributed multiple studies demonstrating FEM-guided placement of wireless sensors at critical blade locations. Their 2018 study presents a comparative evaluation of multiple sensor types — optimised for cost-to-performance ratio — at FEM-identified critical positions, confirming that FEM-guided sensor deployment significantly improves the detection efficiency for fatigue-induced matrix cracking and delamination. Standards bodies including IEC have established certification requirements for blade structural integrity that increasingly reference sensor-based load validation methods.

“FEM-guided sensor deployment significantly improves the detection efficiency for fatigue-induced matrix cracking and delamination — confirming that sensor placement strategy is as important as sensor technology choice.”

Engineering implementations: from root bending moments to delamination detection

The most commercially mature embedded sensing implementation in wind turbine blades is root bending moment monitoring, which directly feeds individual pitch control algorithms to reduce uneven dynamic loads. Vestas Wind Systems’ EP 2022 patent introduces an axial force estimation module that corrects raw load sensor output by accounting for centrifugal axial forces along the blade span — a correction that would otherwise introduce systematic errors into the bending moment estimate during high-speed operation. A companion EP 2021 patent introduces a temperature estimation module that determines in-situ blade temperature from wind turbine operational parameters and applies a thermal compensation correction to the load output, directly addressing thermally induced apparent strain as a known systematic error source in strain-gauge-based load measurement.

For cross-sectional load decomposition, LM Wind Power’s EP 2025 patent introduces a geometric sensor arrangement in the blade transverse plane, with primary and secondary sensor axes oriented relative to the elastic centre point, allowing direct measurement of bending moments about two independent axes at multiple span positions — enabling full 3D load state reconstruction. This complements LM Wind Power’s EP 2018 patent, which uses a sensor set at a non-tip position to estimate bending moments at other longitudinal positions via a processing unit, enabling load extrapolation to locations that are not directly instrumented.

The reliability of blade load sensors is itself a monitoring target. Research from the Technical University of Denmark (2020) presents a fault detection strategy based on residual generators and a Generalized Likelihood Ratio Test for the t-Location Scale distribution, enabling detection of incipient blade load sensor failures that would otherwise propagate undetected into the IPC algorithm, gradually increasing blade loads and fatigue damage. Guidelines from DNV on wind turbine structural integrity increasingly recognise sensor fault propagation as a reliability risk in closed-loop pitch control systems.

Damage detection and delamination surveillance

Beyond continuous load measurement, embedded sensors are employed to detect specific composite damage mechanisms in real time. Siemens Gamesa’s EP 2021 patent on surveillance of delamination propagation uses electrically conductive mechanical fasteners embedded in the blade shell, spar cap, or web, connected in series to form an electrical circuit. Onset or progression of delamination causes one or more fasteners to snap, breaking the circuit and providing an unambiguous binary delamination signal. The same manufacturer’s EP 2023 monitoring system patent takes a dual-use approach: the blade’s existing lightning protection system down conductor is electrically connected to the structural conducting component via equipotential connectors, forming a network of electrical impedances. Monitoring the reception pattern of electrical pulses through this network simultaneously yields health information on both the lightning protection system and the structural component — a notable integration efficiency that reduces additional hardware requirements.

Siemens Gamesa’s EP 2025 patent extends this further by deploying an electrical time-domain reflectometry (TDR) device that measures along first and second electrical paths running the full span of the blade — enabling the spatial localisation of structural changes, such as delamination-induced resistance changes, along the blade length in real time. For a damage-mechanism-oriented approach, the Technical University of Denmark’s 2020 study demonstrates local SHM targeting specific failure types — leading edge erosion, adhesive bond failure, plydrop delamination, and bolt/laminate fatigue — using sensors deployed specifically to detect the physical signature of each mechanism. This mechanism-specific philosophy contrasts with global vibration-based methods and offers higher sensitivity and lower false-alarm rates for known failure modes.

Virtual load sensing and model-fusion architectures

Virtual load sensors — algorithms that reconstruct blade load states from indirect, more reliable measurements fused with physics-based turbine models — offer a cost-effective complement to direct embedded sensing. The primary motivation is the cost and unreliability of physical strain gauges: Goldwind’s 2018 study on a 6 MW wind turbine describes the fusion of generator speed and nacelle acceleration measurements with a high-fidelity Bladed numerical model to estimate structural loads online in real time, explicitly citing strain gauge cost and unreliability as the driver. A similar philosophy is applied by the National Renewable Energy Laboratory (NREL) in a 2020 study that uses an augmented Kalman filter to estimate wind speed, thrust, tower position, and tower loads from a small set of standard SCADA measurements — intended for real-time digital twin applications.

Flanders Make’s 2020 research demonstrates a virtual sensor that estimates incoming planetary stage loads by combining ring-gear surface strain measurements with a physics-based model, deployed for real-time execution on low-cost embedded hardware. The trade-off between algorithm configuration parameters, execution time, and memory footprint is explicitly analysed for embedded deployment constraints — a practical consideration that distinguishes deployable virtual sensors from laboratory demonstrations. Research frameworks from IEEE on cyber-physical systems and digital twins have established the theoretical foundations that underpin these model-fusion architectures.

The ISIS Sensorial Materials Scientific Centre’s 2014 work investigates a hybrid approach for material-integrated SHM in which self-organising mobile multi-agent systems process sensor node outputs at microchip scale, while off-line inverse numerical computation provides spatially resolved load information. This architecture is particularly promising for resource-constrained, distributed blade sensor networks where central processing is impractical. Xi’an Jiaotong University’s 2014 study introduces an improved wavelet denoising algorithm (IRSGWT) and sparse sensor optimisation for CFRP wind turbine blade composites — directly addressing the challenge of minimising sensor count while maintaining spatial damage resolution.

Frontier requirements: energy harvesting, self-calibration, and machine learning

An emerging challenge for embedded sensors in rotating wind turbine blades is power supply — rotating structures cannot easily receive wired power, and battery replacement in sealed composite laminates is impractical over a 20-year blade lifetime. The University of Massachusetts’ WO 2025 patent addresses this directly by integrating a vortex generation-based energy harvester into the blade surface — an air turbine that generates power from the airflow over the blade’s outer surface — to power the embedded sensor system. Makaremi’s US 2013 patent similarly describes integrated power generation for sensor and communication circuitry as part of a comprehensive blade structural integrity system.

Graphic Era University’s IN 2025 patent represents the current state-of-the-art in multi-modal, autonomous blade monitoring architectures: it integrates FBGs, piezoelectric transducers, and acoustic emission sensors with a wireless communication unit, a machine learning data analysis module, and a self-calibration mechanism. The patent explicitly addresses the goal of predictive maintenance and reduction of downtime through real-time anomaly detection — combining the sensor fusion, communication, and intelligence layers that earlier systems treated separately. Research institutions including NREL have similarly identified the integration of machine learning with embedded sensor data streams as a priority for next-generation wind turbine structural monitoring.

The integration of machine learning and data-driven methods with embedded sensor data is an active research frontier. Xi’an Jiaotong University’s work on quantitative damage detection and sparse sensor array optimisation (2014) introduced the IRSGWT wavelet denoising algorithm for CFRP blade composites, demonstrating that algorithmic advances can compensate for sparse physical sensor deployments. ETH Zürich has established vibration-based monitoring benchmarks under varying climate conditions, providing reference datasets against which machine learning models can be validated across seasonal and environmental variation — a prerequisite for reliable field deployment.

“Energy harvesting and autonomous sensor self-calibration are frontier requirements for long-life embedded blade monitoring systems — the blade’s own aerodynamic environment is being leveraged to power the sensors that monitor it.”