Generating Electricity While Walking with Smart Magnetoelastic Insoles

Abstract

Human walking generates renewable and accessible energy with a potential output of up to 67 W. However, effectively harnessing this energy is significantly impeded by the high humidity within the shoe, stemming from both the human body and the surrounding environment. In response to this challenge, we introduce an intrinsically waterproof and biocompatible magnetoelastic smart insole engineered to generate electricity during walking by utilizing the giant magnetoelastic effect. The smart insole features an ultralow internal impedance of ~80 Ω, a substantial current output up to 59.9 mA, and a peak power output of 1.04 mW, performance characteristics that enable the sustainable powering of running lights as well as provide personalized thermoregulation. Furthermore, the insole demonstrates exceptional durability, consistently maintaining peak performance, even with prolonged use and immersion in water. With these extensive advantages, the smart magnetoelastic insole serves as a sustainable, efficient, and robust power source for the Internet of Things, providing a pervasive, decentralized, and mobile energy solution that overcomes the inherent limitations of traditionally centralized and rigidly structured power grids.

Graphical Abstract

The Internet of Things (IoT), utilizing distributed sensors, has fundamentally transformed our daily lives.^1–5 Although each sensor consumes a small amount of power, the cumulative energy demand is immense due to the use of up to trillions of sensors worldwide.^6–8 Traditional electrochemical energy storage devices, such as batteries and capacitors,⁹ are not ideal for powering distributed IoT devices due to their limited lifespans, necessitating frequent replacement. Furthermore, their rigid structures are unsuitable for on-body applications,¹⁰ and the various toxic chemicals they contain pose potential health hazards to users.¹¹ As global energy crises and climate change continue to significantly impede the sustainable progression of human civilization,^12–15 there is a pressing need to develop sustainable energy harvesting methods that are versatile and adaptable to a myriad of applications and environments. Biomechanical motion represents a significant opportunity for energy harvesting applications. Integral to our daily lives,¹⁶ it encompasses essential activities such as heartbeats, respiration, eating, drinking, speaking, and walk- ing.^17–19 Among the diverse biomechanical motions, walking generates one of the highest amounts of mechanical energy, up to 67 W.²⁰ Consequently, harnessing this energy to generate electricity is highly desirable.²¹ Although walking is abundant in nature, the substantial energy it generates typically goes to waste due to the difficulty of efficiently harvesting these small-scale, irregular, and low-frequency signals. By utilizing wearable bioelectronics, we can overcome these challenges, enable continuous energy harvesting, and thereby provide a promising renewable energy source. Previously, various working mechanisms have been employed to harvest energy from human walking, including piezoelectric,²² triboelectric,²³ electromagnetic,²⁴ and electrostatic mechanisms,²⁵ for various applications, including driving electroporation-based disinfection in portable water.²⁶ However, the widespread adoption of these technologies has been hindered by their low current output, high internal impedance, and vulnerability to humidity, which is exacerbated by high humidity levels within shoes due to inadequate heat dissipation and sweat evaporation, as well as exposure to rain and snow.²⁷ While encapsulation strategies can mitigate humidity-related degradation in traditional electromagnetic,^28,29 triboelectric,³⁰ or piezoelectric harvesters, such approaches often introduce additional mechanical stiffness, while the encapsulation layer acts as a mechanical damping layer, reducing the effective transfer of mechanical energy to the active material and thereby decreasing the energy conversion efficiency. These effects further compromise long-term comfort and sensitivity—factors that are particularly critical for wearable insoles subjected to repetitive high-load deformation.²⁶

The magnetoelastic effect, which is the variation of a material’s magnetic field under mechanical stress, is traditionally observed in rigid metals and metal alloys, and has been extensively utilized in civil engineering for building vibration control over the past 156 years.³¹ We recently discovered a giant magnetoelastic effect³² in soft polymer systems in 2021 and leveraged this phenomenon to develop a fundamentally new class of soft magnetoelastic bioelectronics for sensing,^33–35 therapy,^36,37 and energy^38,39 applications. Compared with other human-motion energy-harvesting mechanisms (Table S1), magnetoelastic generators integrate soft magnetized composites and coils into a single monolithic structure, where deformation-induced dipole rearrangement produces magnetic flux variation without requiring rigid-body motion or spatial separation between magnetic and conductive ele- ments.^33,34,40 This compact and all-in-one architecture enables superior structural integrity, reduced thickness, and could improve conformability for footwear applications.⁴¹ Furthermore, humidity tolerance in magnetoelastic generators is not a consequence of their electromagnetic induction principle per se but of their integrated soft-composite architecture. This architecture eliminates exposed sliding interfaces and mechanical gaps that are more challenging to encapsulate in conventional electromagnetic devices, while enabling deformation-induced magnetic particle–particle interaction and magnetic dipole–dipole interaction rearrangement to modulate magnetic flux internally.^42–45 As a result, the magnetoelastic approach enables a soft, monolithic, and moisture-resilient architecture for biomechanical energy harvesting and offers a more wearable and robust solution for repetitive plantar compression in sweat-rich footwear environments where space, comfort, and mechanical durability are critical constraints.

Herein, we introduce the giant magnetoelastic effect in soft polymer systems as a fundamentally new principle for harvesting energy from human walking via a wearable magnetoelastic generator. Specifically, we developed a smart magnetoelastic insole that harnesses the mechanical deformation generated during walking to induce magnetic flux density shifts, thereby enabling electricity generation through electromagnetic induction. Due to the natural ability of magnetic fields to penetrate water, the smart magnetoelastic insole is inherently waterproof, making it a promising solution to mitigate challenges linked to foot humidity caused by both environmental conditions and perspiration from foot confinement. Moreover, the smart insole demonstrates versatile energy harvesting applications, including running light powering and personalized thermoregulation capabilities. The insole’s ability to power running lights, maintain performance even after submersion in water, and endure over 50,000 cycles showcases its potential for continuous, reliable biomechanical energy-harvesting under a wide range of conditions. Additionally, by simulating walking, the insole generated enough electricity to power a fiber heater, raising its temperature by 0.29 °C. This highlights the smart insole’s ability to provide personalized thermoregulation, a crucial feature for maintaining optimal body temperature, especially in varying environmental conditions and for individuals with specific health needs. Given its vast applicability, high current output, exceptional stability, comfort, and inherent waterproof nature, this smart magnetoelastic insole introduces a novel approach to generating electricity while walking. Its sustainable, biocompatible, and noninvasive design shows great potential for applications in thermoregulation and powering running lights, paving the way for widespread adoption as a reliable power source for bioelectronics.⁴⁶

RESULTS AND DISCUSSION

Even under dry conditions, the humidity within shoes is notably elevated due to perspiration and restricted air circulation within the confined space (Figure 1a). Additionally, moisture levels within the shoes can be further increased when walking in wet conditions such as rain or snow. As magnetic fields enable effective water penetration with minimal attenuation, we introduce a waterproof smart magnetoelastic insole for harvesting energy during walking (Figure 1b). As demonstrated in Figure 1c, the smart magnetoelastic insole leverages the giant magnetoelastic effect in an elastomer system featuring a magnetic induction (MI) layer (Figure 1d) and a giant magnetomechanical coupling (MC) layer (Figure 1e) affixed to the insole and subsequently enclosed within a textile layer. The MC layer consists of an elastomer system (Figure S1), which includes 79.3 μm micromagnets dispersed within a silicone matrix (Figures 1f and S2), exhibiting an excellent elasticity of 350% and a Young’s modulus of 804 kPa (Figure S3).

Figure 1.

Smart magnetoelastic insoles for sustainable energy harvesting from human walking. (a) Elevated humidity within the shoe results from sweat accumulation and environmental factors such as rain and snow. (b and c) Schematics illustrating the smart magnetoelastic insole (b) and an enlarged view of its device structure (c). (d and e) The smart magnetoelastic insole includes the magnetic induction (MI) layer (d) and magnetomechanical coupling (MC) layer (e). (f) 3D micro-CT scan of the magnetoelastic elastomer system. Scale bar, 1 mm. (g and h) Schematic illustrating the giant magnetoelastic effect in the elastomer system, showing magnetic flux in the original (g) and compressed (h) states. (i) Magnetic hysteresis loop of the elastomer system under both uniaxial and no stress. (j) Image of the as-fabricated smart magnetoelastic insole submerged in water. Scale bar, 3 cm.

To further optimize the smart insole design, the MI and MC layers were strategically placed beneath the heel to effectively harness biomechanical energy, particularly the weight distribution centered on the hindfoot and forefoot during walking. By combining these two layers, the smart magnetoelastic insole operates on a dual-stage conversion system that transforms mechanical motion into electrical energy. The MC layer, comprising micromagnets evenly distributed within a silicone matrix (Figure 1g), facilitates the conversion of mechanical energy into magnetic flux density shift. This is enabled by the wavy-chain arrangement of micromagnets within the silicone matrix,³² which induces fluctuations in magnetic flux density through magnetic particle–particle and dipole–dipole interactions during deformation (Figure 1h). Further building upon this design, the MI layer, consisting of a copper coil, facilitates the conversion of magnetic flux density shift into electrical energy through magnetic induction:

$$\epsilon =-N\frac{d\varphi }{dt}$$ (1)

where ε, the induced voltage, is determined by the number of turns in the coil, N, the magnetic flux, ϕ, and the time, t. Higher voltage outputs correspond to increased magnetic flux variations and a great number of coil turns. To select the number of turns of coil and provide quantitative justification, we conducted systematic tests on coil configurations with 100, 200, 400, and 600 turns; as the coil number increased, the impedance of the insole during each movement rose proportionally, enhancing signal generation (Figure S4). However, increasing the turn number also elevated the smart insole’s thickness (from 2.8 mm at 100 turns to 7.2 mm at 600 turns), which compromised wearer comfort by introducing bulkiness and reducing flexibility (Figure S5). Balancing these trade-offs, maximizing electrical output while preserving wearability, we selected 400 turns, achieving an optimal current of 70.33 mA at a manageable height (4.1 mm) and high comfort (Table S2). To assess the impact of compression on the magnetic properties of the system, we analyzed its magnetic hysteresis loop before and after applying a preload stress (Figure 1i) to the MC layer, with compression of the system resulting in reductions in both remanence and coercivity. Under dynamic plantar pressure, the MC layer undergoes magnetic flux density variations, generating electricity in the MI layer. Furthermore, we verified the inherent waterproof nature of the smart magnetoelastic insole, which is crucial as insoles typically experience high humidity from perspiration, inadequate circulation, and exposure to environmental moisture like rain and snow. Remarkably, the insole exhibited sustained functionality, even when fully submerged in water (Figure 1j). Moreover, owing to the resilience of the silicone matrix,⁴⁷ this smart magnetoelastic insole not only rebounds effectively from substantial external forces during each step but also boasts biocompatibility, compact size, and waterproofing. These features render it ideal for harvesting energy while walking in a comfortable, convenient, and water-resistant manner.

After assessing the intrinsic waterproof and magnetic characteristics, we examined the giant magnetoelastic effect on the smart insole. The magnetic flux density distribution within the MC layer was evaluated under no applied pressure (Figure 2a) and under an applied force of 300 N (Figure 2b). This layer exhibited a magnetomechanical coupling factor (d₃₃) of 3.19 × 10⁻⁸ T/Pa (Figure 2c), demonstrating its ability to effectively convert biomechanical motion into magnetic flux density shifts. To delve deeper into the scientific basis behind the observed giant magnetoelastic effect within the MC layer, we developed an analytical model using COMSOL Multiphysics. This model integrates solid mechanics and a magnetic field modulus, employing an elastomer system with an assumed Poisson’s ratio of 0.5. The magnetic flux density shift observed at the center of the system, derived from the MC layer and remnant magnetization in the simulation, aligned closely with the experimental observations. Figure 2c illustrates the simulation model’s accurate depiction of the reduction in magnetic field strength as pressure increased in the center of the elastomer system, thus confirming the validity of the model. The simulated magnetic flux density distribution within the vertical direction, depicted in Figure 2d, corresponds with experimental findings, showing a decline as pressure increases from 0 to 800 kPa. Additionally, Figure 2e demonstrates the simulated distribution of magnetic flux density along the vertical axis of the cross-sectional surface (YZ plane, X=0), depicting stress levels ranging from 0 to 800 kPa. The observed reduction in magnetic field strength illustrated in the simulation is consistent with experimental findings. Therefore, the changes in the observed magnetic field in the MC layer under varying pressures stem from the giant magnetoelastic effect in conjunction with the high magnetomechanical coupling factor of 3.19 × 10⁻⁸ T/Pa.

Figure 2.

Giant magnetoelasticity in soft matter and operating principles of the magnetoelastic insole. (a and b) Magnetic flux density distributions of the elastomer system without compression (a) and with compression (b). (c) Magnetic hysteresis loop of the elastomer system under both uniaxial and no stress. (d) Vertical distribution of magnetic flux density in the system under stresses ranging from 0 to 800 kPa. (e) Simulated vertical magnetic flux density distribution on the cross-section surface (YZ plane, X=0) under stresses ranging from 0 to 800 kPa.

Building on this foundation, we further assessed the working mechanism and the smart magnetoelastic insole’s ability to generate electrical signals with every step. The magnetoelastic effect enables the smart insole to harness the two distinct phases of each walking stride: the initial heel strike (phase I, Figure S6a) and the subsequent toe slap with weight transfer (phase II, Figure S6b).⁴⁸ During phase I of each footstep, the MC layer is compressed, inducing a change in the magnetic flux density, which the MI layer then converts into electrical outputs. Conversely, during phase II, as the weight shifts from the heel to the toe, the MC layer relaxes, causing an inverse change in magnetic flux density and, consequently, in output electricity. To characterize the insole’s performance under various conditions, we applied consistent mechanical motions to the insole across frequencies ranging from 0.5 to 3 Hz, evaluating output current (Figure 3a) and voltage (Figure 3b). As the output voltage depends on the change in magnetic flux density over time, higher frequencies increase the rate of change, resulting in shorter peak-to-peak times, higher output voltage, and consequently higher current. Next, the performance of the smart insole was evaluated under varying pressures of 500, 600, and 650 N, which approximate the forces exerted during walking (Figure 3c). As expected, applying higher pressures resulted in increased output voltages caused by the elevated changes in the magnetic flux density. Additionally, as the load increased, the smart insole experienced greater deformation, resulting in longer durations for each compression and release cycle and leading to an increase in peak-to-peak times. Due to the variations in step height during walking, we also measured the output performance across various stepping heights of 10, 20, and 30 cm, as depicted in Figures 3d and S7. As the step height increased, the distance traveled by the foot during each step increased, thereby extending the duration of each cycle and raising peak-to-peak times. The output voltage also rose, aligning consistently with previous measurements, reflecting that larger steps generate greater force due to higher velocity and momentum. The performance of the smart insole was evaluated under various movements to assess its voltage (Figure 3e) and current (Figure 3f) outputs during walking, running, and jumping. As the movements progressed from walking to running to jumping, the frequency of steps increased, resulting in shorter peak-to-peak times. Consequently, the rise in step frequency and impact force across these movements led to higher current and voltage output from walking to running to jumping.⁴⁹ The smart magnetoelastic insole, despite its compact size, achieved a peak open-circuit voltage of 558 mV and a maximum current of 59.9 mA during jumping. To ensure wear comfort, prevent odor, and allow moisture evaporation during prolonged wear, we tested user thermal comfort, water vapor transmission rate, and air permeability. For thermal comfort assessment, the textilefacing side was placed in direct contact with a heated plate at 35 °C (simulating skin temperature), while the magnetic side was exposed to ambient air. Steady-state temperatures stabilized at 33.5 °C on the textile surface and 30.2 °C on the magnetic side (Figure 3g), yielding a low thermal resistance and evaporative resistance, which effectively dissipated heat and moisture without inducing discomfort. Water vapor transmission rate testing demonstrated excellent moisture management (Figure 3h). Air permeability averaged 1516 L/m²/s at 49 Pa and 2353 L/m²/s at a 100 Pa pressure differential (Figure 3i), reflecting robust ventilation through the porous textile layer while maintaining structural integrity under foot pressure. These properties collectively affirm the insole’s suitability for extended daily use, balancing energy harvesting with breathability.

Figure 3.

The electric output of the smart magnetoelastic insole. (a) Output current (blue) and peak-to-peak time interval (red) under identical applied forces at varying frequencies. An increase in frequency led to an increase in current, along with a reduction in peak-to-peak time. (b) Output voltage (blue) and peak-to-peak time interval (red) under identical applied forces at varying frequencies. An increase in frequency led to an increase in voltage, along with a reduction in peak-to-peak time. (c) Output voltage (blue) and peak-to-peak time interval (red) under different applied forces representative of those typically applied to the human body. As the applied force increased, there was a corresponding rise in both peak-to-peak time and voltage. (d) Peak-to-peak time interval (red) and output voltage (blue) as functions of step height. An increase in step height resulted in a greater peak-to-peak time and voltage output. (e) Output voltage (blue) and peak-to-peak time interval (red) during different human movements: walking, running, and jumping. Across these movements, the voltage increased while peak-to-peak time decreased. (f) Output current (blue) and peak-to-peak time interval (red) during different human movements: walking, running, and jumping. Across these movements, current increased, while peak-to-peak time decreased. (g) Thermal comfort evaluation of the smart magnetoelastic insole under simulated skin conditions (35 °C and 50% RH). (h) Water vapor transmission rate of the smart magnetoelastic insole over 12 h at 37 °C and 50% RH. (i) Air permeability of the smart magnetoelastic insole under different air pressures.

Following the standard characterization, we further evaluated the insole’s energy-harvesting capabilities. We first obtained the output voltage and instantaneous power across a range of resistances applied to the smart insole. Figure 4a demonstrates that the output voltage rose progressively as the applied resistance increased from 1 Ω to 10 kΩ. Despite a compact device size, peak power was achieved at a resistance of 80 Ω, reaching 1.04 mW. In addition to the significant impact of power output on the performance of the smart insole, long-term stability is a critical consideration given their intended continuous wear. Consequently, after assessing power output, our focus shifted to evaluating device durability by subjecting the smart magnetoelastic insole to a constant force at a consistent frequency with a shaker, simulating approximately 10 days of human walking (50,000 cycles, assuming 5000 steps per day). The insole demonstrated no obvious performance degradation even after more than 50,000 cycles (Figure 4b).

Figure 4.

Demonstration of the smart magnetoelastic insole as a wearable and sustainable power source. (a) Voltage (blue) and power (red) characteristics across various resistances. The output voltage progressively increased with higher applied resistance, while peak power, reaching 1.04 mW, was achieved at a resistance of approximately 80 Ω. (b) Long-term stability of the smart magnetoelastic insole over more than 50,000 cycles, with the inset demonstrating consistent current outputs after 20,050 cycles. (c) Charging 22, 47, and 100 μF capacitors by continuously stepping on the smart magnetoelastic insole. (d) Stability comparison of the smart magnetoelastic insole before and after submersion in water. Submersion in water showed no significant impacts on the peak-to-peak time (red) or output current (blue). (e) Maximum short-circuit current (blue) and open-circuit voltage (red) of the smart magnetoelastic insole before and after immersion in PBS for 48h. Postimmersion, current and voltage retention exceeded 90% of preimmersion values, demonstrating robust electrical stability under moisture exposure. PBS simulates actual sweating conditions, ensuring reliability in real-world wear. (f) Output current of the smart magnetoelastic insole before (blue) and after (red) immersion in PBS during different human movements: walking, running, and jumping. (g) Output current of the smart magnetoelastic insole during different human movements: walking, running, and jumping with different numbers of coil turns. (h) Capability of the smart insole to illuminate more than 40 LEDs postsubmersion in water. Scale bar: 5 cm. (i) Activation of running lights from a single step using the smart magnetoelastic insole. Scale bar: 3 cm. (j) Sample voltage output generated from walking using the smart magnetoelastic insole.

To further highlight the potential of the smart magnetoelastic insole as a sustainable power source for wearable devices, we demonstrated its energy harvesting capability by charging three different capacitors (Figure S8). Within 120 s, the insole successfully charged 22, 47, and 100 μF capacitors to 1.2, 0.6, and 0.3 V, respectively, as shown in Figure 4c. Due to the high amount of perspiration produced by the foot, the intrinsic waterproofness of the smart magnetoelastic insole is vital to its functionality and must be properly validated. To achieve this, the insole was completely submerged in water, demonstrating no performance degradation, even in wet environments (Figure 4d). Furthermore, the phosphate-buffered saline (PBS) solution was used to simulate actual sweating or raining conditions. Postimmersion, maximum short-circuit current and open-circuit voltage retention exceeded 90% of preimmersion values, demonstrating robust electrical stability under moisture exposure (Figure 4e). During different movements (jumping, running, and walking), the output current of the smart insole shows good electrical stability before and after PBS immersion, as shown in Figure 4f. We further examined the durability of coil configurations (100, 200, 400, and 600 turns) under prolonged moisture exposure (Figure 4g) by evaluating peak output current across varied movements (walking, running, and jumping) after 48 h immersion in PBS. All designs retained >90% of preimmersion current postsoak, with the 400-turn optimal configuration showing the least degradation and consistent performance across movements. After submersion, the smart insole effectively powered over 40 light-emitting diodes (LEDs) using the minimal biomechanical energy generated from hand tapping (Figure 4h and Video S1). Therefore, the smart magnetoelastic insole is well-suited for daily use, as it can endure the demanding conditions associated with everyday activities. Additionally, we evaluated the real-time performance of the device by connecting four LEDs to the insole and securely attaching them to the shoe to be used as running lights. Throughout the utilization of the insole, each step prompted illumination of the running lights (Figure 4i). Furthermore, a continuous walking pattern generated output voltages consistent with those illustrated in Figure 4j, with the running lights methodically ignited with each step (Video S2). Hence, the smart magnetoelastic insole can sustainably power running lights, offering a multitude of advantages, such as increased runner visibility during night runs and enhanced safety. These results showcase the smart magnetoelastic insole’s ability to consistently harvest biomechanical energy even under harsh conditions.

In addition to powering running lights, the energy harvested during walking can be utilized for personalized thermoregulation.⁵⁰ Personalized thermoregulation is crucial for maintaining optimal body temperature, especially under varying environmental conditions and for individuals with specific health requirements. The smart insole enables personalized thermoregulation and represents a significant advancement in wearable technology, enhancing comfort and benefiting users by adapting to their individual thermal needs. To evaluate the thermoregulation performance, we used an electrical shaker to simulate walking steps. This device continuously deformed the smart insole (Figure 5a), closely mimicking the mechanical stresses experienced during walking. At the core of the design is a fiber heater specifically designed to match the resistance of the MI layer embedded within the insole. The fiber heater was seamlessly integrated into the smart insole, ensuring optimal wearing comfort and efficient performance (Figure 5b). To facilitate the flow of electricity generated by the insole, conductive wires were used to establish a robust connection between the fiber heater and the MI layer (Figure 5c). During simulated walking, the smart insole was subjected to continuous deformation, leading to effective electricity generation (Figure 5d). As illustrated in Figure 5e, neither the smart magnetoelastic insole nor a pure magnetic film induced any adverse skin reactions after 1 week of continuous wear, thereby confirming its excellent biocompatibility. To test how device performance varies with user weight, gait variability, or walking speed, we recruited healthy volunteers of 50, 60, and 80 kg and performed standardized trials on a treadmill. Figure 5f demonstrates that the output current rose progressively as the weights of the volunteers increased. Next, the performance of the smart insole was evaluated under different walk speeds (0.33, 0.5, 1, and 2 Hz), which approximate the speed exerted during walking (Figure S9). Output current remained remarkably stable during walking trials across all weight groups, underscoring the device’s reliability and adaptability for diverse users in prolonged real-world scenarios (Figure 5g).

Figure 5.

Demonstration of the smart magnetoelastic insole for active personalized thermoregulation. (a) Experimental setup for the personalized thermoregulation experiment with an enlarged view illustrating the position of the smart insole within the shoes. Scale bar: 3 cm. (b and c) Photos demonstrating the integration of the fiber heater into the smart insole (b), and an enlarged view of the fiber heater which is seamlessly integrated within the smart insole (c). Scale bars: 8 mm (b) and 2.5 mm (c). (d) Voltage outputs generated by simulated steps applied to the smart insole, with the blue band highlighting the effective output value required to drive the fiber heater. (e) Photograph showing the skin irritation results of a piece of the smart magnetoelastic insole and an impermeable magnetic film on the forearm. Scale bar: 1.0 cm. (f) Output current of the smart magnetoelastic insole under different gaits (walking, jumping, and running) for body weights of 50 kg (blue), 60 kg (red), and 80 kg (gray). As body weight increased, peak current rose proportionally. (g) Sample current output generated from jumping for body weight of 50 kg using the smart magnetoelastic insole. (h) Thermoregulation performance measurement of the fiber heater powered by the smart insole, shown with an enlarged infrared photo which depicts the elevated temperature.

This electrical energy was then utilized to increase the temperature of the fiber heater through Joule heating. Within 5 min of continuous simulated walking, the temperature of the fiber heater rose by 0.29 °C (Figure 5h). After ceasing the simulated walking, the temperature quickly reverted to near baseline levels, indicating that the increase in temperature was due to the fiber heater rather than the ambient environment. Although modest, this subtle temperature elevation (0.3–2 °C) has been shown to provide meaningful physiological benefits, including enhanced foot microcirculation, reduced cold stress, and improved perceived thermal comfort, particularly in cool environments or for individuals with cold-sensitive feet.^51,52 For patients with diabetes or peripheral circulation impairments, even small increases in foot sole temperature can help maintain vascular health, mitigate risks of ulceration, and counteract evaporative cooling from sweat.^53–55 By effectively harnessing the energy generated from walking, the smart insole can dynamically adjust the thermal environment around the foot, enhancing user comfort in cold conditions. The integration of the fiber heater into the smart insole not only demonstrates the feasibility of energy harvesting for thermoregulation but also paves the way for further innovations in wearable technology.

CONCLUSIONS

In conclusion, in this work, we introduce a smart magnetoelastic insole capable of serving as an inherently waterproof and biocompatible means to sustainably harvest biomechanical energy from walking, which exhibited an exceptionally low internal impedance of approximately 80 Ω along with a short-circuit current up to 59.9 mA. The compact smart magnetoelastic insole achieved a peak power output of 1.04 mW when tested under an 80 Ω load. When inserted into a shoe, the insole successfully powered running lights with just a single step, highlighting its capability as a sustainable power source and acting as a promising method to enhance runner safety during night runs. Additionally, by harnessing mechanical energy, the smart insole was able to effectively increase the temperature of a fiber heater by 0.29 °C, demonstrating the insole’s promise for personalized thermoregulation. The smart magnetoelastic insole overcomes the limitations of traditional biomechanical motion energy harvesters, such as low current output and high internal impedance by combining its inherent waterproofness with its high flexibility, wearing comfort, and stable output performance. The smart magnetoelastic insole emerges as a compelling solution for harvesting energy from walking, thus functioning as a sustainable and versatile power solution for bioelectronics.

Methods

Fabrication of the Smart Magnetoelastic Insoles.

Ecoflex 00–30 elastomer (smooth-on) was combined with micromagnets. The mixture underwent thorough stirring to achieve a uniform distribution of the micromagnets throughout the elastomer system. Subsequently, the uncured composite was transferred into customized molds and cured in an oven (ThermoFisher) at 60 °C for 3 h to solidify. The elastomer system was then magnetized by using an impulse magnetizer (IM-10–30, ASC Scientific) to ensure remnant magnetization within the system. Following magnetization, the elastomer system was inserted into a copper coil encased in a fabric. This fabric was then sewn into the hindfoot region of the insole to form the smart magnetoelastic insole.

Smart Magnetoelastic Insole.

Scanning electron microscopy (SEM) using a Zeiss Supra 40VP instrument was employed for structural characterization of the elastomer system. A digital Gaussmeter (TD8620, Tunkia) was utilized to measure the magnetic flux density shift. Magnetic hysteresis loops were analyzed by using a SQUID magnetometer (MPMS3, Quantum Design). 3D micro-CT images were scanned by using a μCT scanner (CrumpCAT).

Electrical Performance Characterization of Smart Magnetoelastic Insoles.

The voltage output of the smart magnetoelastic insole was tested using a programmable electrometer, Keithley 6514, and a Stanford low-noise preamplifier (model SR560). Current output was measured with a Stanford low-noise-current preamplifier (model SR570). The insole was secured on a vibration shaker (Labworks ET-015) to apply external pressure. Furthermore, continuous sinusoidal signals were generated, driven by a power amplifier (PA-1200) and a function generator (SIGLENT SDG 2122X). The smart insole was directly connected to a diode bridge rectifier (MBSK16SE), facilitating the conversion of alternating current into direct current. Voltage signals were amplified using a toroidal transformer amplifier to charge capacitors and illuminate LEDs.

Thermal Comfort and Breathability Characterization of Smart Magnetoelastic Insoles.

Thermal comfort was evaluated by placing the textile-facing side of a 30 × 30 mm sample against a 35 °C heated plate to simulate human skin temperature, while the magnetic-facing surface was exposed to ambient laboratory conditions (22–24 °C, 40–50% RH). Surface temperatures were recorded using a thermometer (DANOPLUS DP-373) until steady state. Water vapor transmission rate was measured using the ASTM E96 BW protocol at 23 °C and 50% RH. Air permeability was quantified according to ISO-9237 using an air permeability tester (TX-726) under differential pressures of 49 and 100 Pa.

Personalized Thermoregulation.

The fiber heater, incorporating an 80 Ω axial-lead resistor, was integrated into the smart insole. To simulate walking steps at a frequency of 20 Hz, vibration shakers (Labworks ET-015), power amplifiers (PA-1200), and customized 3D-printed rods were employed. Temperature measurements were taken using a high-precision temperature sensor (TMP117), controlled by custom Arduino code, with the sensor node positioned close to the surface of the fiber heater during testing. Ambient temperature fluctuations were minimized by using a foam box enclosure.

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsnano.5c20508.

Movies S1: Smart insole lights up LEDs after submersion in water (MP4)

Movies S2: Smart insole harvests the walking energy to power running lights (MP4)

Additional details regarding the microstructure of the magnetoelastic elastomer, size distribution of micromagnets, stress–strain curves and Young’s modulus, coil impedances, photographs of coils with different turns, device–foot interaction analysis, output current under various stepping conditions, circuit design for walking energy harvesting, and body-weight-dependent performance (Figures S1–S9, Tables S1–S2, and Note S1) (PDF)