Research Progress of Self-Powered Gait Monitoring Sensor Based on Triboelectric Nanogenerator

Abstract

In recent years, technologies in the field of gait monitoring, such as gait parameter analysis, health monitoring, and medical diagnosis, have become increasingly mature. Gait monitoring technology has emerged as an effective means for disease prevention and diagnosis. Triboelectric nanogenerator technology not only overcomes the limitations of relying on external power sources and frequent battery replacements but also offers advantages such as low cost, lightweight, a wide range of material options, and ease of manufacturing. This review introduces the common working modes of triboelectric nanogenerators and summarizes recent advances in self-powered gait monitoring applications (e.g., gait analysis, fall detection, rehabilitation assessment, and identity recognition), and highlights persistent challenges such as wearability, washability of fabric-based devices, reliability, system integration, and miniaturization, along with proposed solutions.

1. Introduction

With the advent of the technological era, intelligent information technology has been widely applied in multiple fields [1,2,3,4,5,6,7]. Real-time and accurate gait monitoring shows enormous application potential, such as medical diagnosis, rehabilitation assessment and treatment, and health management [8,9,10,11,12,13]. Existing traditional gait monitoring methods include visual observation and measurement. The analysis results of visual observation are somewhat subjective, and are related to the experience and skills of doctors and physical therapists. The measurement method can objectively provide the kinematic parameters, kinetic parameters, and electromyographic activity parameters of the gait, but it cannot display data in real time. With the advancement of science and technology, high-precision and objective gait monitoring technologies have enabled real-time data collection and analysis, and have been applied in the fields of healthcare, rehabilitation, and sports training [14,15,16,17,18]. However, they still have shortcomings, such as reliance on external power sources, strict operating conditions, and environmental pollution issues and economic burdens brought about by frequent battery replacements. To address these deficiencies, many researchers have begun to pay attention to triboelectric nanogenerators (TENGs). TENGs harvest energy from the surrounding environment through the coupling effect of triboelectricity and electrostatic induction and convert it into electrical energy, such as wind energy, mechanical energy, and droplet energy [19,20,21,22,23,24,25,26]. Consequently, they serve various fields, including health monitoring [27,28,29,30], motion monitoring [31,32,33], energy harvesting [34,35,36], and self-powered sensors [37,38,39]. TENGs are regarded as promising technologies due to their low cost, simple manufacturing process, wide range of material choices, strong portability, and high energy conversion efficiency [40,41,42]. Many studies have successfully integrated TENGs into practical scenarios, such as insoles, clothing and running tracks for real-time health and motion monitoring, and signal analysis [43,44,45,46,47,48,49,50,51,52]. They have achieved outstanding research results and practical application value.

Here, the progress in the field of TENGs for self-powered gait monitoring is examined. First, a comprehensive overview of the four common working modes of TENG technology is provided, including detailed explanations of two predominant modes: vertical contact-separation (C-S) mode and single-electrode (S-E) mode. Second, the research progress of self-powered gait monitoring sensors driven by TENGs in the fields of basic gait information monitoring, abnormal gait detection and fall monitoring, and rehabilitation assessment and assisted training, as well as identity recognition, is systematically summarized (Figure 1).

2. Triboelectric Nanogenerator

2.1. Working Mechanism of Triboelectric Nanogenerator

The working principle of the TENG is that the coupled effect of the triboelectric effect and electrostatic induction converts biomechanical energy into electrical energy [53]. In essence, the triboelectric effect refers to the phenomenon that when materials with different electronegativities come into contact, electron transfer occurs, resulting in both materials carrying equal amounts of opposite charges [54]. Electrostatic induction occurs when a charged object approaches a conductor. The attractive and repulsive electrostatic forces among charges induce the redistribution of charges inside the conductor [55]. During the contact process of two materials with different electronegativities, electron transfer takes place due to the triboelectric effect. After the two materials are separated, electrostatic induction generates an induced potential difference between the materials and the electrodes, which in turn prompts the flow of electrons between the external circuit and the electrodes [56,57,58,59,60]. Therefore, during the periodic contact and separation of materials with different electronegativities, the TENG produces corresponding alternating current signals.

2.2. Working Modes of Triboelectric Nanogenerator

There are four basic working modes of a TENG: vertical contact-separation (C-S) mode, lateral-sliding (L-S) mode, single-electrode mode (S-E), and freestanding-layer (F-E) mode (Figure 2a). This review mainly introduces the common vertical contact-separation mode and single-electrode mode. The vertical contact-separation mode is the most fundamental mode of the TENG. It is composed of two friction layers with different electronegativities and electrodes stacked, where the electrodes are located at the opposite ends of the contact surfaces of the two friction layers. Under the action of an external force, the two friction layers approach each other and electron exchange occurs, which causes the two friction layers to carry equal amounts of opposite charges, respectively. When the two friction layers separate, the equal amounts of opposite charges between them form a potential difference. If the two electrodes are connected through an external load, electrons flow from the high-potential electrode to the low-potential electrode to achieve charge balance. When an external force is applied to the friction layers again, the transferred charges return to the original electrodes and generate a reverse current. When the friction layers repeatedly make contact and separate, the TENG can generate periodic alternating current (Figure 2(bi)).

In the single-electrode mode, the friction layer consists of a freely moving object and a friction-layer material with different electronegativity. Compared with the vertical contact-separation mode, it has one fewer electrode (Figure 2(bii)). When the contact surfaces of the moving object and the friction layer approach or move away from each other, the local electric-field distribution can be changed, creating a potential difference between the electrode and the ground wire, thus driving the flow of electrons and generating current output. The simplification of the structure in this mode has advantages for harvesting mechanical energy in certain specific situations. However, due to the limitation of the electrostatic shielding effect, the electrical output of the single-electrode is weak, making it unsuitable for application scenarios that require a relatively strong electrical output.

3. Applications of Self-Powered Triboelectric Nanogenerator in Gait Monitoring

3.1. Basic Gait Information Monitoring

Human gait parameter information includes gait cycle, kinematic parameters, spatio-temporal parameters, kinetic parameters, etc. These parameters can objectively reflect an individual’s walking characteristics. In recent years, an increasing number of researchers of TENGs have shown great interest in self-powered gait monitoring sensors. Among them, some researchers have achieved gait monitoring by integrating TENGs into insoles or shoe soles. For example, Zhao et al. fabricated a TENG based on electrospun composite nanofibers to construct a self-powered gait analysis system. The system can monitor walking speed and acceleration and realize data visualization [61]. There are three TENGs in this system, and the placement of the sensors is determined through plantar pressure simulation analysis (Figure 3(ai)). Two of the TENGs form a sensing module and are placed at the forefoot and heel positions, and one TENG is used as a charging module and placed at the heel position (Figure 3(aii)). Gait monitoring is achieved by relying on the two TENGs in the sensing module to collect the electrical signals generated during walking. The original output signals are processed by a charge amplifier circuit and then converted into sensing signals readable by the microcontroller unit (MCU). The processed signals present as transistor–transistor logic (TTL) signals (Figure 3(aiii)). The two TENGs used for obtaining signals can acquire stable signals, and the TENG of the charging module can drive the MCU to collect and calculate data. Based on the above mechanism, Zhang et al. developed an application program that can not only record the gait parameters of the exerciser, but also realize the visualization of gait parameters (Figure 3(aiv)).

Placing TENGs on the soles of the feet for signal collection poses greater requirements for the mechanical properties and electrical output performance of the TENG materials. To date, many researchers have focused on developing triboelectric materials with mechanical properties and device structures to improve performance output. For instance, Zhao et al. proposed an origami-shaped metamaterial [62]. This metamaterial can modulate its mechanical properties through geometric design modification. Subsequently, it was fabricated into a shoe sole via 4D printing technology. This kind of shoe sole is a TENG with self-power and self-sensing capabilities. The TENG operates based on the vertical contact-separation working mode. In terms of its structure selection, the structure of three topological units is used as the substrate (Figure 3(bi)). Aluminum foil is pasted on the dark-colored part, which serves as the conductive layer and the positive triboelectric layer. On the opposite part, the aluminum foil acts as the electrode, and the piezoelectric and flexible thin film (PFF) is fixed above the aluminum foil as the negative triboelectric layer. Zhao et al. integrated the TENG into the shoe sole to monitor the voltage signals generated during walking in real time, and analyzed the foot pressure situation through the signal output. In addition, voltage signals generated by slow walking, fast walking, jogging, and running are different. Different walking patterns can be identified by analyzing the signal types, thus enabling gait pattern monitoring (Figure 3(bii)). Shao et al. fabricated an integrated TENG with numerous pores for gait cycle monitoring [63]. It is worth noting that, compared with the same type of TENG without a pore structure, its electrical output performance is increased by approximately 10 times. The porous TENG operates in a single-electrode working mode. Each individual pore can be regarded as a separate TENG (Figure 3(ci)). Its principle of electricity generation lies in the periodic contact-separation between the carbon nanofibers (CNFs) in the pores and the rubber. The porous structure of this TENG ensures a high degree of sensitivity to pressure and deformation (Figure 3(cii)). Moreover, Shao et al. placed the porous TENG in the shoe sole. Different gait patterns result in different signal outputs, which can reflect that toe-walking, marching, and stomping generate different output signals (Figure 3(ciii)). Thus, it is possible to determine the movement postures and analyze the gait cycle.

In addition to integrating TENGs into shoe soles, some researchers have also integrated them with clothing. Zhu et al. invented a self-powered cotton sock based on a hybrid piezoelectric and triboelectric mechanism [64]. Compared with integration into insoles and soles, socks with sensing functions can monitor gait states more conveniently and for a longer time. In Zhu et al.’s study, a textile TENG was fabricated using PEDOT: PSS-coated fabric. The TENG was used as part of an integrated piezoelectric sensor (Figure 3(di)). To better obtain triboelectric signals during contact, Zhu et al. constructed an intelligent passage using polytetrafluoroethylene (PTFE) and aluminum electrodes. When the participants walked along the intelligent passage while wearing cotton socks, the signal output at different stages of the gait cycle could be obtained (Figure 3(dii)). After conducting tests on different participants, the researchers verified the stability of the system in gait cycle signal detection applications (Figure 3(diii)).

All the above-mentioned studies involve contact-type TENGs. Non-contact TENGs also have enormous application potential in the field of gait monitoring. For example, Xi et al. proposed a non-contact TENG for human motion monitoring, which can be used for speed monitoring and motion tracking [65]. This TENG operates in a single-electrode mode. The human body serves as the electronegative triboelectric layer of the TENG, while printing paper and metal film act as the positive triboelectric layer and the electrode. Figure 3(ei) depicts the charge transfer pattern followed when the right leg moves. Two TENGs are fixed on the wall as speed sensors (Figure 3(eii)). When a person walks or runs along different paths, Zhao et al. achieved real-time monitoring of human movement speed by analyzing the voltage peaks and the passing time (Figure 3(eiii)). In addition, the non-contact TENG can also serve as a human motion tracking sensor. Even if there is a partition wall or an obstacle between the human body and the TENG, the sensor can still track the human body’s movement trajectory based on the signal differences generated by different movement paths (Figure 3(eiv)). In summary, the self-powered gait monitoring sensors based on TENGs have been widely applied in the field of sports monitoring. The durability of the device, stability of signal output, and the detection sensitivity are crucial for long-term and high-precision gait monitoring.

3.2. Abnormal Gait Detection and Fall Monitoring

During the process of human walking gait, long-term out-toeing and in-toeing in the gait may cause damage to the human body. Therefore, the detection of abnormal gaits, such as out-toeing and in-toeing, is of great significance for human health monitoring. Currently, numerous TENG-based pressure sensors are used for abnormal gait monitoring. Seong et al. fabricated a TENG with synergistic complementarity and installed six arrays of complementary nanopattern triboelectric nanogenerators (CN-TENGs) on the sole of the shoe to construct a gait monitoring sensing system for diagnosing the out-toeing and in-toeing in the gait (Figure 4(ai,aii)) [66]. When the sole exhibits in-toeing, the electrical signal generated on the inner side is greater than that on the outer side. Conversely, when the sole exhibits out-toeing, the electrical signal on the outer side of the sole is greater than that on the inner side (Figure 4(aiii)). Similarly, Hu et al. developed a self-powered sensing insole to identify abnormal gait by dynamically capturing the changes in plantar stress [67]. The insole contains four TENG sensing units based on the single-electrode mode. The structure of the sensors and their distribution in the insole are shown in Figure 3(bi). Hu et al. used a 16-channel signal acquisition system to diagnose six common abnormal gaits (Figure 3(bii,biii)).

In addition, TENG-based strain sensors have shown great potential for abnormal gait monitoring. Wang et al. fabricated a hydrogel-based strain TENG using graphene oxide and polyacrylamide hydrogel (GO-PAM) [68]. Its output power is 2.2 times that of pure PAM hydrogel. Wang et al. integrated the strain TENG into an insole for abnormal gait detection. As shown in Figure 4(ci), the sensors on the sole of the foot acquire signals that are converted and then transmitted from the MCU to a personal computer. Subsequently, the raw signals undergo a fast Fourier transform. Finally, the ANN, decision tree, and random forest models are constructed to recognize and analyze the gaits. Wang et al. successfully applied three machine learning models to identify three types of pathological gaits (Figure 4(cii)).

Most of the existing studies on TENGs for gait monitoring are mainly based on the interactions between solid–solid, solid–liquid, and liquid–liquid. The mechanism of these studies is the contact and separation of two independent interfaces, and this mechanism is vulnerable to the interference of the surrounding environment. Xiong et al. invented a TENG that utilized the interaction between gas and solid to achieve electron transfer [69]. Its working mechanism was realized through the coupling of the triboelectric effect and electrostatic induction, as well as the dynamic migration of charges in the air (Figure 4(di)). Compared with the dense polysiloxane-dimethylglyoxime-based polyurethane (PDPU), the porous PDPU has better structural self-healing ability, surface adhesiveness, and output stability (Figure 4(dii)). It is integrated into shoes to obtain the motion signals of human walking (Figure 4(div)), which can analyze the normal and dragging gait conditions of the human body, and it also has the self-healing ability (Figure 4(diii)). The technology of knitting yarns has also been widely applied in wearable gait monitoring sensing technology. Ahmed et al. invented a wearable fabric TENG for fall monitoring [70]. The fabric is designed in a 3D diamond pattern to enhance the surface interaction at the contact interface. This fabric TENG also features strong breathability, washability, and durability. Ahmed et al. placed two TENGs at the positions of the forefoot and the heel of the insole, respectively, and constructed an abnormal gait detection and fall monitoring system (Figure 4(ei)). This system is capable of detecting abnormal gait in patients in real time. Figure 4(eii) shows that patients unconsciously accelerate their walking speed after deviating from their normal stride pattern. At the end of the gait, although they have the intention to move, their feet are unable to move, and characteristic signals appear immediately. When a patient falls, negative signals are generated, and no positive signals are produced until the patient stands up (Figure 4(eiii)); this is important for safety monitoring of older people at home. For instance, when elderly individuals experience a fall and remain incapacitated on the ground, the corresponding signal remains unchanged. Conversely, if the elderly person manages to rise following the fall, a transformation in the signal occurs. By leveraging this differential signal behavior, it can systematically assess and decide whether emergency assistance should be promptly summoned. The abnormal gait detection and fall monitoring system constructed based on this TENG demonstrates its application potential in the field of health monitoring.

In addition, Wang et al. developed a textile TENG based on polyvinylidene difluoride (PVDF) and polyamide 66 (PA66) nanofiber coaxial yarns (Figure 4(fi)) [71]. The textile is realized through a coaxial conjugate electrospinning process. During the textile process, a novel conductive gel composed of liquid metal, silver nanowires (AgNWs), and polyvinyl alcohol (PVA) is filled between the conductive wire core and the nanofiber shell. This filling technology is used to address the poor contact between the core and the shell, and, meanwhile, endows the fabric with stronger electrical conductivity and antibacterial properties. Inspired by the research on the TENG woven with this yarn, Wang et al. developed an insole based on plain fabric for gait monitoring [71]. The insole contains a total of 15 pressure sensors (Figure 4(fii)). As the plantar pressure distributions of the out-toeing gait and the in-toeing gait are different, they are abnormal compared with the standard gait signals. Based on this principle, it can be used for the diagnosis of abnormal gaits and the assessment of rehabilitation (Figure 4(fiii)). The application of TENG technology in the field of gait monitoring provides data references for the diagnosis of abnormal gaits and fall monitoring, and solves the problem of achieving accurate analysis of abnormal gaits relying on complex diagnostic instruments.

3.3. Rehabilitation Assessment and Auxiliary Training

The advent of the intelligent era of medical devices has propelled the rapid development of scientific and technological medical diagnosis and evaluation. Currently, an increasing number of lower limb rehabilitation assessment and auxiliary functional training devices based on TENG technology have been reported. Inspired by the skin of sharks, Cheng et al. invented a TENG based on shark skin bionics [72]. It uses liquid metal and Ecoflex as the friction layers (Figure 5(aii)). A gait monitoring system constructed based on the bionic TENG, together with the sensors of four TENGs installed on the soles of the feet and the data acquisition system, jointly constructs a gait analysis system (Figure 5(ai)). Cheng et al. divided the gait cycle generated during human walking into four stages (Figure 5(aiii)), and the gait of the participants was analyzed based on the signals generated in these four stages. Notably, the states of the iliopsoas muscle and the tibialis anterior muscle were monitored in real time to assess injury rehabilitation (Figure 5(aiv)). The gait monitoring system can also be used to monitor anterior cruciate ligament (ACL) tears and meniscus injuries (Figure 5(av)). Zhang et al. also developed a wearable electronic sensor based on a TENG and applied it to a robot-assisted rehabilitation system (Figure 5(bi)) [73]. The workflow of the robot-assisted rehabilitation system is shown in Figure 5(bii). The robot-assisted rehabilitation system not only includes the intelligent insole for analyzing the gait, but also four intelligent sensing seat belts for monitoring the movement of the waist (Figure 5(biii)), and the system is able to provide patients with individualized training programs that are tailored to their condition (Figure 5(biv)).

The TENG based on fabric can be combined with clothing and accessories to realize the monitoring of gait signals. The research by Wei et al. introduced a gait monitoring system of a self-powered multipoint body motion sensing network with an all-textile structure, as shown in Figure 5c [74]. The system relies on machine learning technology to analyze the limb movement cycle signals and dynamic indicators in real time, achieving precise monitoring of the user’s gait. Meanwhile, they also designed a personalized auxiliary rehabilitation training system (Figure 5d). In addition, the system can also customize personalized and targeted rehabilitation plans for patients. Figure 5e demonstrates a human–machine interaction platform for patient rehabilitation. In addition, some researchers have developed a TENG sensor that integrates the functions of assessment and correction for pathological gaits. Xu et al. integrated a TENG using a polymer with self-healing and shape memory functions [75]. This polymer-based TENG exhibited excellent output stability and self-healing ability, that is, when the device is broken, the output performance can still recover to the original state after healing (Figure 5(fi)). The healable and shape-memory dual functional polymers triboelectric nanogenerator (HSP-TENG) also had a large shape fixation rate and a shape recovery rate in high-temperature environments (Figure 5(fii)). The contact and separation between the foot and the corrective insole based on the HSP-TENG generates signals. The information provided by these signals can not only enable early detection of biomechanical anomalies associated with flatfoot, but the design of the sensor’s shape can also offer certain arch support for patients with flatfoot (Figure 5(fiii)).

The research by Guo et al. proposed a self-powered wearable multidimensional motion sensor that can sense vertical acceleration and planar angular velocity [76]. The sensor is composed of an acceleration sensor and an angular velocity sensor (Figure 5(gi)). The researchers integrated the sensor into a belt to construct a VR-assisted training system, and the workflow of this system is shown in Figure 5(gii). In the study, the multidimensional motion sensor for monitoring movement was also placed at the ankle position, and the kicking motion was identified through 1D convolutional neural network (CNN) deep learning technology to achieve the control of VR games (Figure 5(giii)). The VR game controlled by this multidimensional motion sensor has the potential to provide a novel auxiliary training approach for the personalized training of patients with abnormal lower limb gaits.

3.4. Machine Learning-Assisted Identity Recognition

In recent years, regarding the applied research of TENGs in the field of gait monitoring, apart from monitoring basic gaits, abnormal gaits, and providing auxiliary training, the identity recognition system based on TENG sensing technology has also received extensive attention. The system is realized by integrating the TENG sensing technology with machine learning technology. The most common form is to place the TENG sensor in the fabric to obtain signals and then use various algorithms to identify these signals. Zhang et al. designed a TENG with a wearable PVA/acrylic fluorescent layer and developed a triboelectric sensor array that simulates the structure of nerve cell protrusions that achieved effective perception of multidirectional stress [77]. It consists of a composite PVA/acrylic fluorescent layer and a pair of electrodes. One of the electrodes is coated with a PTFE film to generate induced negative charges (Figure 6(ai)). Due to its excellent flexibility, the TENG is integrated into the insole in the form of a 10-sensor array by researchers for gait monitoring (Figure 6(aii)). The insole-based STA can obtain personal information composed of triboelectric signals. By combining it with a deep learning CNN model, an intelligent system for individual identification was constructed. Figure 6(aiii) shows the triboelectric signals generated by different participants while walking. Through the training and learning of signals among different individuals by the CNN model, individual identification can be completed accurately. Its accuracy rate can reach 99.75% (Figure 6(aiv)). In addition, Xu et al. combined the invented triboelectric sensor with an insole to create an intelligent insole (Figure 6(bi)). The sensor material is composed of Galinstan and silicone rubber. Galinstan is used as the electrode and Ecoflex is used as the negative friction layer [78]. Machine learning technology can identify the subtle differences in triboelectric signals and, thus, reveal information in depth. The study combined the machine learning algorithm of CNN with the intelligent sensing insole to identify individual information. First, the action datasets generated by different individuals are collected. Then, the calculation is carried out through the CNN model. Finally, the signals generated by them can be identified. Furthermore, the researchers optimized the CNN model by evaluating the cross-entropy loss function. By setting two convolutional layers and two pooling layers, the accuracy rate of individual identification ultimately reaches 94.86% (Figure 6(bii)).

In addition to combining sensors with insoles, some scholars have also explored other methods and successfully integrated them with the human body. Han et al. invented a flexible and stretchable ring-shaped TENG. By attaching it to the calf, a gait recognition device was constructed [79]. The ring-shaped TENG is made of a rubber tube filled with physiological saline (Figure 6(ci)). When the wearer’s muscles contract and relax, the contact state between the skin and the rubber changes, creating a potential difference between the device’s electrodes and the ground, thereby enabling the flow of electrons (Figure 6(cii)). The ring-shaped TENG has strong stretchability and is suitable for being worn on multiple parts of the human body. In their study, the sensor was worn on the calf to capture electrical signals, and then identity recognition was carried out based on these electrical signals (Figure 6(ciii)). Li et al. proposed a triboelectric gait sensing system integrated into a carpet for human activity recognition [80]. The sensor uses two materials, copper and PTFE, as the friction layers to achieve the triboelectric effect. The researchers designed the sensor in the shape of an eight-channel array and integrated it into the carpet. When the participants walk on the carpet, real-time electrical signals can be generated. The individual information is identified according to the characteristics of the gait signals of different participants (Figure 6(di)). In terms of machine learning technology, researchers proposed a Residual Dense-BiLSTM deep learning model evolved from the classic long short-term memory (LSTM) algorithm (Figure 6(dii)). The accuracy recognition rate of this model was 99.4% (Figure 6(diii)). The identity recognition technology based on TENGs is becoming increasingly mature, providing technical references and theoretical guidance for more convenient and intelligent identity recognition systems. In general, the identity recognition achieved by the gait sensors based on TENGs mainly relies on the sensors to obtain signals and machine learning technology to identify these signals. Among them, the CNN machine learning technology, with its advantages such as high efficiency in feature extraction and strong robustness, has become an important tool and technical means in the field of gait monitoring.

3.5. Comparison Between TENGs Based on Different Materials and Structure

In the field of gait monitoring, TENGs with various structures and designs each present a unique set of advantages and disadvantages. TENGs integrated into insoles or soles can monitor multiple parameters and sensitively analyze gait cycles, but they require high material performance and are complicated to install and maintain. TENGs integrated into socks are easy to wear and enable continuous and stable gait monitoring, but there are deficiencies in signal acquisition and mechanical support. Furthermore, non-contact TENGs break through the contact limitation and simplify the design of the system, but they make it difficult to acquire fine gait parameters and are susceptible to environmental interference. Other TENGs with special structural design, such as those based on hydrogels, TENGs with self-healing properties and TENGs made of specific fabrics, have unique advantages, such as adjustable performance and high-power output, but they generally face the problems of difficult research and development, complicated preparation processes, limited application scenarios, or insufficient stability. The distinct characteristics of various TENG structural configurations, material compositions, and design architectures in terms of output performance and application suitability are summarized in

Table 1. Comparison between TENGs based on different materials and structure.

Structure	Working Mode	Tribo-Materials	TENG Output	Detection Things	Refs.
Nanofiber	C-S	PVDF/BaTiO3/MWCNT and silver fabric	~374 V, ~10.2 mA	Step frequency, step speed and acceleration monitoring.	[61]
Polymer	TENG + PENG	Al and PFF	~131 V, ~1.1 μA	Motion status and step frequency monitoring.	[62]
Elastomer	S-E	Rubber and carbon nanofiber	~91 V, ~2.87 μA	Abnormal gait detection.	[63]
Textile	TENG + PENG	PEDOT: PSS coated fabrics and PTFE	~1.71 mW	Gait pattern recognition, motion tracking.	[64]
Non-contact	S-E	Body and paper	~2 V	Motion speed tracking.	[65]
Copolymer	C-S	Cu and Teflon	~442 V, ~22.2 μA	Abnormal gait monitoring.	[66]
Elastomer	Gas–solid interaction	PDPU porous elastomer and air	~5.2 V	Parkinsonian gait recognition.	[69]
Textile	C-S	Nylon/cotton yarn and PDMS/BaTiO3	~763 V, ~20.4 μA	Abnormal gait monitoring fall monitoring.	[70]
Yarn-woven	C-S	PVDF and PA66 yarns	~3.45 V, ~62.1 mW	Abnormal gait detection and joint movement monitoring.	[71]
Bionic	S-E	Ecoflex and liquid metals	~32 V, ~19.04 mW	Detection of neuromuscular disorders and rehabilitation self-assessment.	[72]
Textile	C-S	Ecoflex and Ni-fabric	~20 V, ~1.3 μW	User recognition. Rehabilitation monitoring.	[73]
Textile	C-S	PE sheet yarn and conductive fabric	~9 V	Biological gait recognition, assisted rehabilitation training.	[74]
Polymer	S-E	PBA polymer and movable object		Flatfoot treatment. Gait analysis.	[75]
Triboelectric band	S-E	Rubber tube and skin	~89.4 V	Identity recognition. Step frequency and speed monitoring.	[79]
Stacked structures	C-S	PTFE and copper foils	~80 V, ~10.58 μW	Identity recognition. Gait recognition. Fitness monitoring.	[80]

.

4. Conclusions and Prospects

The advent of the IoT era has accelerated the intelligent and digital development of human body monitoring research. In the field of gait monitoring, the combination of sensing technology with wireless signal technology and machine learning technology has enabled reliable, convenient, and efficient gait monitoring. This paper summarizes the research progress over the past decade in the fields of basic gait information monitoring, abnormal gait detection, and fall monitoring; rehabilitation assessment and auxiliary training; and machine learning-assisted identity recognition based on TENG technology. However, in order to further explore its potential value and expand its application scope, it is necessary to further improve the comprehensive performance of the TENG and address the existing limitations.

(1)

Barriers to high integration of TENG sensors with health monitoring platforms. The integration of TENG-based gait sensors with health monitoring platforms faces several challenges for real-world applications. For example, the gait characteristics of different individuals vary greatly, and the TENG-based gait monitoring sensors may not be able to accurately capture and recognize the gait characteristics of all individuals when collecting data. If the sensor cannot adaptively adjust to these differences, the data collected may be inaccurate, affecting the subsequent analysis and evaluation on the health monitoring platform. Moreover, environmental factors, such as temperature, humidity, and electromagnetic interference, can degrade sensor performance, introducing data deviations and misinterpreting health status. Additionally, compatibility issues arise between TENGs’ signal outputs and existing platforms. Due to their unique signal characteristics, additional circuitry or algorithms are often required for data standardization, increasing system integration complexity.

(2)

Comfort of the sensor. During the process of gait diagnosis for patients or long-term monitoring, the discomfort experienced by users when wearing the sensor may affect the accuracy of the signals. The discomfort may lead to skin irritation or other physical issues for the users. Therefore, to ensure optimal comfort during prolonged wear, gait monitoring sensors require comprehensive optimization of their size, shape, and material properties, along with improved wearing methods. During design, the sensor dimensions should be precisely determined through ergonomic analysis of foot biomechanics across diverse populations, measuring both force distribution and range of motion during gait. The shape must conform to natural foot contours to minimize movement interference, while material selection should prioritize soft, skin-friendly biocompatible substrates. Furthermore, the wearing system should incorporate user-friendly fixation mechanisms that maintain sensor stability without restricting movement or causing discomfort, thereby enhancing both functionality and wearability.

(3)

Washability of fabric-based TENGs. Most of the sensors used for gait monitoring are wearable. Their hygienic conditions need to be taken into consideration, especially those sensors placed on the soles of the feet. Bacteria that breed from sweat or other secretions secreted during exercise not only pose a threat to the safety of human skin but also make it unsuitable for long-term wearing. In view of this, researchers should explore textile materials of TENGs with washability and self-cleaning to create hygienic conditions for their contact with human skin. For instance, by implementing functionalized antimicrobial embedding strategies, antimicrobial agents are incorporated into triboelectric polymer matrices. This approach enables the device to exhibit self-cleaning properties, effectively reducing the growth of bacteria and maintaining a hygienic environment. Additionally, through optimizing the structural design in the preparation process, the device’s wash resistance can be enhanced. It might involve modifying the material’s composition ratios, adjusting the manufacturing process parameters, or adopting novel fabrication techniques to ensure that the device can withstand repeated washing cycles without significant degradation of its performance or structure.

(4)

Stability and durability of the sensor. As a sensor for gait monitoring, the TENG is usually integrated with socks or insoles, or directly contacts the human skin. Gait monitoring is achieved during physical movement, so the impact of chemical substances in human sweat on the sensor cannot be ignored. These impacts may reduce the stability of the sensor. Moreover, environmental factors such as humidity and temperature can potentially compromise the sensor’s performance stability. To solve the above practical problems, on one hand, it is essential to actively research and develop new anti-interference materials. These include corrosion-resistant electrodes, triboelectric polymers resistant to sweat and environmental changes, and materials stable in temperature and humidity. On the other hand, optimizing device encapsulation is crucial. This involves selecting waterproof, moisture-proof, and heat-insulating polymers for tight wrapping, designing a multilayer structure, and configuring ventilation holes rationally. This protects against external substances and balances pressure. Through these efforts, the stability and durability of TENG sensors are likely to be significantly enhanced, potentially promoting broader application in gait monitoring.

(5)

System integration and miniaturization. Currently, many gait diagnosis systems and auxiliary training systems based on TENG sensors are composed of multiple modules, which inevitably leads to a relatively large volume. A large volume may limit the application conditions. Therefore, achieving the functional integration of the system and the miniaturization of the device are the keys to the widespread application of the sensing system. For system integration, standardizing interfaces and communication protocols across modules is fundamental. Additionally, meticulous optimization of module design is necessary. Regarding equipment miniaturization, leveraging advanced manufacturing techniques, such as microelectromechanical systems (MEMS) and precision machining, holds great promise. These technologies allow the fabrication of high-precision components within confined spaces, enabling a higher degree of component integration. Through optimized spatial layouts and innovative multifunctional integration designs, multiple functional elements can be consolidated into a compact unit. This approach is likely to reduce the physical footprint of the equipment while maintaining its core functionality, potentially facilitating the broader adoption of TENG-based sensing systems across various applications.