A Systematic Review of Insole Sensor Technology: Recent Studies and Future Directions

Abstract

Background: Integrating diverse sensor technologies into smart insoles offers significant potential for monitoring biomechanical metrics; enhancing sports performance; and managing therapeutic interventions, diseases, disorders, and other health-related issues. The variation in sensor types and applications requires a systematic review to synthesize existing evidence and guide future innovations. Objectives: This review aims to identify, categorize, and critically evaluate the various sensors used in smart insoles, focusing on their technical specifications, application scopes, and validity. Methods: Following the PRISMA guidelines, a search was conducted in three major electronic databases, namely, PubMed, Scopus, and Web of Science, for relevant literature published from 2014 to 2024. Other works not located in the mentioned databases were added manually by parallel searches on related themes and suggestions from the website of the databases. To be eligible, studies were required to describe sensor implementation in insoles, specify the sensor types, and report on either validation experiments or practical outcomes. Results: The search identified 33 qualifying studies. Proper analysis revealed a dominance of pressure sensors, with accelerometers and gyroscopes also being widely used. Critical applications included gait analysis, posture correction, and real-time athletic and rehabilitation feedback. The review also examined the relative effectiveness of different sensor configurations. Conclusions: This systematic review comprehensively classifies sensor technologies within smart insoles and highlights their broad application potential across various fields. Future research should aim to standardize measurement protocols, enhance sensor integration, and advance data processing techniques to boost functionality and clinical applicability.

1. Introduction

The use of sensor technology in wearable devices, specifically in the field of footwear, has attracted considerable interest in both the improvement of performance and the rehabilitation of individuals [1,2,3]. Insoles integrated with sensors serve as a vital convergence of biomechanics, data science, and sports science, providing a distinctive platform for real-time monitoring of biomechanical factors and physiological responses. These devices are crucial for athletes seeking to enhance performance and avoid injuries [4], as well as those in rehabilitation, as they offer valuable information that informs therapeutic actions [5].

Accurate numerical information about the movement and forces involved in foot mechanics is crucial in multiple areas of scientific investigation. Therefore, assessing plantar forces is an essential technique for examining human body movement [6]. In general, the trade-off between the flexibility of wearable devices and the precision of fixed systems must be considered. Force plates are widely recognized as the most reliable way for precisely identifying gait events. Nevertheless, force plates are mostly utilized in controlled laboratory environments, and the limited availability of these plates restricts the number of steps that may be accurately measured. On the other hand, sensor insoles provide a convenient method for gathering data in both controlled and real-world settings, with few limitations on their use. They offer exceptional efficiency, adaptability, and portability [6].

Advancements in technology have resulted in a rise in sensor technologies that may be integrated into insoles, with each sensor providing distinct data types and valuable information. These include pressure sensors that measure weight distribution and gait patterns, as well as accelerometers and gyroscopes that detect movement dynamics and orientation. The wide range of sensor technologies accessible offers a vast array of possibilities for investigation, but it also creates difficulty in identifying which sensors provide the most dependable and valuable data for specific applications. Exploring alternative combinations and types of sensors is necessary due to the crucial considerations of the number, location, and applicability of sensors in the insole.

This systematic review seeks to synthesize and gather existing information on this issue, offering a complete study of the many types of sensors used in insoles to improve sports performance and aid in rehabilitation. Hence, this research aims to methodically identify and assess the many types of sensors presently utilized in insoles, evaluating their applications, efficacy, and the data quality they offer. The primary inquiry is as follows: “What types of sensors are employed in insoles to enhance performance and facilitate rehabilitation?” This inquiry seeks to clarify the predominant sensor technologies, emphasizing their advantages and limitations in these two important situations. This comprehensive analysis aims to provide organized and informed perspectives that aid in choosing the most suitable sensor technology for specific requirements.

2. Materials and Methods

The primary objective of this study was to examine relevant literature focusing on research articles, review articles, and systematic reviews published in English that are aligned with the research objectives. This investigation targeted publications from 2014 onwards, ensuring that the review reflected the most recent developments and contributions in the field.

Three distinct databases were chosen for the literature search: PubMed, Scopus, and Web of Science (WoS). Each one of these databases offers unique strengths that are aligned with the research needs. PubMed, known for its extensive collection of biomedical and clinical studies, was chosen for its relevance and depth of clinical research. This database is renowned for its comprehensive coverage of topics related to healthcare, medicine, and related fields, making it an invaluable resource for the current study.

Scopus and WoS were selected for their robust combination of quantitative and qualitative studies, along with their expansive coverage of scientific disciplines. Scopus is recognized for its extensive indexing of peer-reviewed literature, offering a wide range of scientific articles across diverse disciplines. Likewise, WoS also provides a comprehensive indexing of multidisciplinary research.

Utilizing these three databases ensured a wide-ranging overview of the relevant literature, enhancing our analysis’s quality and scope. Moreover, the databases chosen provided a balanced representation of clinical and broader scientific research, which was crucial for achieving the study’s objectives. Nevertheless, to complete the information provided by the search in the three databases, searches were conducted without following guidelines, resulting in exploring articles suggested online or included in the works found previously. The proposed methodology allowed us to capture various perspectives and insights, contributing to a more nuanced understanding of the topics under investigation.

In summary, the study’s methodological approach facilitated a thorough and balanced literature review, leveraging each database’s unique strengths to provide a robust and comprehensive analysis.

A literature search was performed in the Web of Science (WoS) database using the query: ((“wearable sensor” OR “wearable sensors”) AND (“insole” OR “insoles”) AND “biomechanics”),

Table 1. Keyword combinations used for searches in the selected databases.

Database Name	Platform	Date Coverage	Date of Search	Search Within	Source Title	Command	# of Results
Scopus	Elsevier	2014–2024	30 April 2024	Article Title, Abstract, Keywords	All	“wearable sensor” OR “wearable sensors” AND “insole” OR “insoles” AND “biomechanics”	51
Scopus	Elsevier	2014–2024	30 April 2024	Article Title, Abstract, Keywords	All	“wearable sensor” OR “wearable sensors” AND “insole” OR “insoles”	51
Scopus	Elsevier	2019–2024	30 April 2024	Article Title, Abstract, Keywords	All	“wearable sensor” OR “wearable sensors” AND “insole” OR “insoles” AND “biomechanics”	41
Scopus	Elsevier	2019–2024	30 April 2024	Article Title, Abstract, Keywords	All	“wearable sensor” OR “wearable sensors” AND “insole” OR “insoles”	41
Web of Science	Clarivate	–2024	30 April 2024	All Fields	All	“wearable sensor” OR “wearable sensors” AND “insole” OR “insoles” AND “biomechanics”	3508 *
Web of Science	Clarivate	2006–2024	30 April 2024	Article Title, Abstract, Author Keywords	All	“wearable sensor” OR “wearable sensors” AND “insole” OR “insoles” AND “biomechanics”	105
Web of Science	Clarivate	2014–2024	30 April 2024	Article Title, Abstract, Author Keywords	All	“wearable sensor” OR “wearable sensors” AND “insole” OR “insoles” AND “biomechanics”	90
Web of Science	Clarivate	2019–2024	30 April 2024	Article Title, Abstract, Author Keywords	All	“wearable sensor” OR “wearable sensors” AND “insole” OR “insoles” AND “biomechanics”	68
PubMed	NLB/NIH	2014–2024	2 May 2024	All Fields	All	“wearable sensor” OR “wearable sensors” AND “smart insole” OR “smart insoles” AND “insole” AND “performance” AND “rehabilitation”	3
PubMed	NLB/NIH	2014–2024	2 May 2024	All Fields	All	“wearable sensor” AND “insole” AND (“performance” OR “rehabilitation”)	7 **
Web of Science	Clarivate	Until 2024	2 May 2024	Article Title, Abstract, Author Keywords	All	“wearable sensor” OR “wearable sensors” AND “smart insole” OR “smart insoles” OR “insole” AND “performance” AND “rehabilitation”	88
Web of Science	Clarivate	Until 2024	2 May 2024	All Fields	All	“wearable sensor” AND “insole” AND (“performance” OR “rehabilitation”)	19 **
Scopus	Elsevier	Until 2024	2 May 2024	Article Title, Abstract, Keywords	All	“wearable sensor” OR “wearable sensors” AND “smart insole” OR “smart insoles” AND “insole” AND “performance” AND “rehabilitation”	4
Scopus	Elsevier	Until 2024	2 May 2024	Keywords	All	“wearable sensor” AND “insole” AND (“performance” OR “rehabilitation”)	16 **

* Number of results analyzed to produce the network visualization graphic (Figure 1). ** Number of results analyzed to obtain the information in all bar charts (Figure 2, Figure 3 and Figure 4).

. The search produced a total of 3508 results. Afterwards, the data was evaluated using VOS viewer software, version 1.6.20, Netherlands, with a specific focus on titles, abstracts, and keywords as presented in Figure 1, wherein the clustering of specific phrases resulting from the search is illustrated. To enhance the accuracy of this search, the terms “wearable sensor”, “insole”, “performance”, and “rehabilitation” were employed as keywords. These keywords were selected based on their alignment with the primary purpose and research topic. Preliminary searches allowed us to optimize the final query. Hence, this iterative procedure improved the search technique, ensuring the proper connection to our research. The analysis conducted with VOS viewer yielded insights into the thematic clustering of the chosen terms, hence strengthening the significance of the selected keywords in this study. The inclusion of “performance” and “rehabilitation” as supplementary keywords was of utmost importance, as they directly pertained to the proposed utilization of wearable sensor technology in biomechanical investigations. The intention was to encompass pertinent research that addresses both performance-enhancing and rehabilitation applications, ensuring a full overview of the literature in both areas. The methodological approach showcases a structured and focused search strategy, utilizing sophisticated bibliometric methods to extract pertinent literature. The meticulous procedure of iteratively refining the search terms and employing VOS viewer for thematic analysis highlights the thoroughness and profundity of our literature study, which attempted to thoroughly investigate the research subject.

According to Figure 2, from 2014 until 2023, the number of records related to the searched keywords increased. However, it is noticeable that they augmented especially significantly between 2018 and 2021. It is possible that the manifested behavior appeared because of COVID pandemic, which was associated with worsening mobility and physical function even if hospitalization was not required to treat the virus. This upward trend also suggests a growing interest or advancements in the field.

As expected, visualizing the data from Figure 3, it is possible to see that most records available in this area/field of study (studies) are essentially research articles.

Upon analyzing Figure 4, it becomes apparent that most studies on this topic have been published within the domain of electrical engineering. The interconnectivity of different disciplines is evident, resulting in a wide array of outcomes. This analysis indicates that the objectives described in the current study are not adequately covered and are spread out across several fields of study. The data provided and evaluated in Figure 2, Figure 3 and Figure 4 were exclusively gathered from the Web of Science database using the first command. This constraint emphasizes the necessity for a more all-encompassing strategy to more effectively correspond with the goals of this project and offer a unified comprehension of the topic.

This study followed the PRISMA [7] guidelines for systematic reviews. Fifty-one works were part of the review; most of them (42) were retrieved from the three databases, and a total of nine works were added manually. Figure 5 presents show the PRISMA flow diagram, illustrating the screening procedure followed. The choice of the selected databases ensured that low-quality and low-impact articles were not considered. In this review, the focus of the study was directly on the types of sensors used in insoles (or shoe insoles) for performance and rehabilitation.

The PRISMA flow diagram lists other major reasons for exclusion (Figure 5). From the total number of works searched, eleven duplicate records were excluded, as well as two records that were not related to the activities of study, one report that was not available for consultation, and finally four conference papers because the initial decision was not to include these kinds of works.

Figure 6 provides a detailed analysis of the dataset used in this investigation, showing the number and specific categories of investigated scholarly works. The dataset is primarily composed of articles which make up the majority of the sources evaluated. Nevertheless, it was deemed necessary to include additional forms of literature to ensure a thorough examination. As a result, the dataset was expanded by including seven review articles and one systematic review. Incorporating a wider range of source types is essential to encompass a more comprehensive array of academic views and perspectives, thus enhancing the strength and thoroughness of the study’s conclusions.

3. Results

3.1. Overview of Research Articles on “Sensorized” Insoles

The results of this study are presented in

Table 2. Summary of the focus of each work, the equipment utilized, the types of sensors studied, the number of sensors in the insole, and the option of the wireless parameter considered for this systematic review.

Study/Work	Year of Publication	Type of Document	Focus of Study	Equipment	Types of Sensors	Number of Sensors in Insole	Wireless (Yes/No)	Number of Articles Included
[8]	2022	Research article	Monitoring of forces applied to the tibial bone	Shoes and insoles	Inertial measurement unit (IMU) and pressure sensors	ND *	Y	49
[9]	2022	Research article	Validation of pressure-sensing insoles in patients with Parkinson’s disease	Insoles	IMU and pressure sensors	6-axis IMU; 18	Y	42
[10]	2021	Research article	Gait analysis using smart insoles in elderly people and Parkinson patients	Insoles	IMU and pressure sensors	6-axis IMU; 16	Y	54
[11]	2022	Research article	Real-time gait event detection/foot angular kinematics with IMU	Plastic plates	IMU and pressure sensors	tri-axial accelerometer, tri-axial gyroscope, and tri-axial magnetometer; Pedar-X, Novel, DE with 99 each	N	31
[12]	2023	Research article	Estimation and analysis of human gait	Insoles	IMUs and pressure sensors	Sensor node; 3-axis accelerometer and gyroscope; 8 each	N	13
[13]	2016	Research article	Analysis of gait in fallers and non-fallers in the elderly population	Bands, belt, and insoles	IMUs and pressure sensors	Tri-axial accelerometers at the head, pelvis, and left and right shanks; pressure-sensing insoles (F-Scan 3000E, Tekscan, Boston, MA, USA) with 954 each	N	46
[14]	2021	Research article	Collecting and monitoring plantar pressure	Insoles	Piezoresistive sensors	6 textile piezoresistive sensors (TPRS) each	ND *	38
[15]	2020	Review article	Gait phase detection in lower limb prostheses	ND *	IMU, pressure sensors, ground reaction force (GRF) sensor, angular sensors, and force sensors (FSs)	Force-sensing resistors (FSRs), force sensors such as footswitches and foot-pressure insoles	Y/N	98
[16]	2017	Review article	Detection and monitoring gait disorders	Insoles	IMU and pressure sensors	1–64	Y/N	105
[5]	2022	Review article	Plantar-pressure-based insole gait monitoring techniques for disease monitoring and analysis	Insoles	IMU and pressure sensors	4–16	Y/N	229
[1]	2024	Review article	Intelligent wearable systems in health and sports	Camera, smartwatch, IMU smart ring, smart shirt, smart insole	IMUs (accelerometer, gyroscope), magnetometer, MIMU (magneto-inertial measurement unit), bioelectric sensors, biometric sensors, environmental sensors, optical and chemical sensors, flexible sensors	ND *	Y/N	296
[17]	2016	Research article	Real-time human foot motion	Shoe	IMUs (accelerometer, gyroscope, and magnetometer)	3	N	40
[18]	2023	Review article	Wearable technology in sports	Wearable technology (WT)	Physiological, environmental, biomechanical, and location sensors	ND *	Y/N	191
[3]	2022	Review article	Smart insoles	Insole	Accelerometer, gyroscope, IMU, pressure sensors, piezoresistive sensors, capacitive sensors, temperature sensor	ND *	Y/N	175
[6]	2017	Research article	Validation of Moticon’s OpenGo sensor insoles during gait, jumps, balance, and cross-country skiing	OpenGo sensor insole system, PedarX sensor insole, AMTI force-plate systems, and neutral shoes	Capacitive sensors	16 pressure sensors and 6-axis IMU; 99 pressure sensors	Y/N	23
[2]	2022	Review article	Wearables for biomechanical performance optimization and risk assessment	WT	Exoskeletons, IMU, pressure sensors, surface electromyography (EMG)	ND *	Y/N	73
[19]	2022	Research article	A conductive bamboo fabric with controllable resistance for tailoring wearable sensors	Sleeves and insoles	Strain and pressure sensors	4 each	N	28
[20]	2022	Research article	A new 3D, microfluidic-oriented, multi-functional, and highly stretchable soft wearable sensor	Insole, bands (for the wrist, elbow, and knee)	3D soft sensors (resistive sensors)	2	N	51
[21]	2021	Research article	A shoe-integrated sensor system for pressure evaluation	Shoe and insole	FSR sensors	5 FSR sensors and 8 pressure sensors	Y	17
[22]	2023	Research article	Automatic body-segment and side recognition of an IMU sensor during gait	Bands, force plates, insole	IMUs (tri-axial angular velocity and acceleration sensors) and pressure sensors (capacitive)	Pedar-X, Novel, DE with 99 each	N	44
[23]	2022	Research article	Customized textile capacitive insole sensor for center-of-pressure analysis	Insole	Textile capacitive sensor (pressure sensors)	10 each	N	29
[24]	2015	Research article	Estimation of ground reaction forces and ankle moment	Orthopedic shoe with an insole, force plates	Tissue force sensor (pressure sensors)	8 each	N	29
[25]	2023	Research article	Evaluation of hemiplegic gait based on plantar pressure and inertial sensors	Insole	IMUs (gyroscope, magnetometer, accelerometer) and pressure sensors	7 IMUs and 2 pressure insoles	N	38
[26]	2016	Research article	Kinematic analysis of human gait based on wearable sensor system for gait rehabilitation	Shoe and insole	FSRs and accelerometer	7 FSRs, triaxial accelerometer	N	36
[27]	2022	Research article	Open-set user identification using gait pattern analysis based on an ensemble deep neural network	Insole	IMUs and pressure sensors	FootLogger (8 pressure sensors, 3D-axis accelerometer, and 3D-axis gyroscope).	Y	52
[28]	2017	Research article	Prospective fall-risk prediction models for older adults	Shoe, insole, bands, and belt	IMUs and pressure sensors	Tri-axial accelerometers at the head, pelvis, and left and right shanks; Pressure-sensing insoles (F-Scan 3000E, Tekscan, Boston, MA, USA) with 954 each	N	44
[29]	2016	Research article	Smart insole and gait monitoring in daily life	Insole	IMUs and pressure sensors	Textile-based pressure sensors (48) and IMUs (3-axis accelerometer, 3-axis gyroscope, and 3-axis magnetometer)	Y	25
[30]	2014	Research article	SmartStep: a fully integrated, low-power insole monitor	Shoe and insole	IMU and pressure sensors	3D accelerometer, 3D gyroscope, and 3 FSR402 pressure sensors	Y	35
[31]	2021	Research article	Textile-film sensors for a comfortable intelligent pressure-sensing insole	Foam insoles with piezoresistive films (pressure-sensing insoles)	Pressure sensors	4 pressure sensors	N	35
[32]	2018	Research article	Wearable sensor system for detecting gait parameters of abnormal gaits	Smart shoes	Range sensors, force sensors, and inertial sensors	4 range sensors (VL53L0x, STM, Geneva, Switzerland), 6 force sensors, inertial sensors (MPU9250, InvenSense, USA) each	Y	36
[33]	2021	Systematic review	Wearable-sensor-based real-time gait detection	WT	IMU, insole pressure sensors, EMG sensors	ND *	Y/N	138

* ND—not defined.

. The classification of works includes eight columns encompassing the author’s name, the article’s nature, the research’s main focus, the equipment used within the study, the type and number of sensors, and whether they use wireless technology.

According to the information provided in the table above (Table 2), we present in the following figure (Figure 7) some insole-based systems that were utilized in various works. We chose four commercial designs because there was more information available regarding that equipment, while there was little information about some of the prototypes and the designs elaborated on the works referred previously. The dates mentioned in the following figure represent the year that each work was produced and published.

3.2. Sensor Technologies Available on the Market

From the classification presented in Section 3.1, it was possible to identify the available equipment on the market, which was employed in several studies. Hence,

Table 3. Summary of some equipment available on the market and the specifications (type of sensors, number of sensors, presence or absence of a wireless component, and application(s)) of each one.

Device Name	Company/Enterprise	Type(s) of Sensor(s)	Number of Sensors in Insole	Wireless (Yes/No)	Application(s)
Moticon SCIENCE	Moticon (München, Germany)	Pressure sensors, accelerometers, and gyroscopes	16 pressure sensors and 6-axis IMU	Y	Motion analysis for clinical research and sports science
FootLogger	3L Labs (Chicago, IL, USA)	Pressure sensors	(8 pressure sensors, 3D-axis accelerometer, and 3D-axis gyroscope).	Y	Monitoring foot health for rehabilitation and injury prevention
Digitsole	Digitsole (Houston, TX, USA)	IMU	2 detachable connected sensors	Y	Mobility assessment, follow-up/monitoring on site or remotely, rehabilitation follow-up
Nurvv Run	Nurvv (London, UK)	Pressure sensors	16 each	N	Cadence, step length, foot strike
F-Scan	Tekscan (Norwood, MA, USA)	Pressure sensors	954 each	N	Gait analysis parameters such as pressure and force data
Parotec-System	Paromed (Neubeuern, Germany)	Pressure sensors (Piezoresistive)	24	N	Gait studies
Pedar system	Novel (Munich, Germany)	Pressure sensors (capacitive)	99	N	Gait studies
BioFoot	IBV (Valencia, Spain)	Pressure sensors (piezoelectric)	64	Y	Gait analysis parameters such as pressure
OpenGo	Moticon (München, Germany)	Pressure sensors, accelerometers, and temperature sensors	16 pressure sensors and 6-axis IMU	Y	Gait studies, activity tracking
Real-Time Rehab system	Veristride (South Salt Lake, UT, USA)	Pressure sensors	ND *	Y	Gait studies (rehabilitation—amputees)
SurroSense Rx (wrist and feet)	Orpyx Medical Technologies (Calgary, Canada)	Pressure sensors	8 each	Y	Monitoring foot health for rehabilitation and injury prevention (for example, foot ulcers)
Kinematix Tune	Kinematix (North Vancouver, Canada)	Pressure sensor and accelerometer	ND *	Y	Feet symmetry analysis regarding ground-contact time, heel contact, foot strike; combines feet information with GPS parameters: speed, distance
IDM Perform	JUMPSTARTCSR (North Vancouver, Canada)	Accelerometer and gyroscope	ND *	N	Fitness, medical, and industrial

* ND—non-defined.

makes the classification of the several devices.

4. Insole Sensor Technology

4.1. Pressure Sensors

Pressure sensors are electronic devices specifically engineered to quantify and transform pressure into electrical impulses. They are frequently utilized to measure pressure in the hands and feet, facilitating the examination of user movement. Pressure sensors integrated into intelligent insoles offer significant data by assessing the pressure applied during walking activities [3,10]. In addition, these sensors can measure the pressure distribution across the foot. This is crucial for determining the center of pressure (CoP) as well as other gait characteristics, including step count, gait cycle duration, swing duration, and foot-ground interaction events [33]. Recent developments suggest that innovative materials can precisely and sensitively collect data about the detection of standing posture, identification of sports shoes, and activity modes [3]. However, the most precise technique for measuring CoP is utilizing a force plate integrated into the ground. This makes it possible to capture the immediate ground reaction force while maintaining a stationary stance [23].

Chen et al. [5], in their work, compare various pressure-sensing mechanisms and state the advantages and disadvantages of each one. The authors demonstrated that the piezoresistive and capacitive sensors are low-cost and have a simple structure and fabrication process, but both have difficulties regarding temperature. They also stated that piezoresistive, capacitive, and piezoelectric sensors are more appropriate for monitoring and preventing neurological and orthopedic diseases. Whereas piezoresistive sensors present a problem concerning power consumption, capacitive sensors and piezoelectric have lower power consumption. On the other hand, piezoresistive and piezoelectric sensors support multi-axis pressure sensing. Piezoelectric sensors are more affordable and have higher force sensitivity but cannot detect static pressure.

4.1.1. Capacitive

Capacitive sensors quantify the magnitude of force by detecting variations in the separation between the upper and lower electrodes within an elastomer medium. This technique offers a superior degree of force sensitivity and exceptional dynamic performance. However, the range of force the sensor can detect is limited, and it is susceptible to humidity and electromagnetic interference (EMI). A capacitive pressure sensor comprises two primary layers: the dielectric layer and the lead-out electrode layer [5]. These sensors offer significant benefits because they can be used with biological materials, opening new opportunities for monitoring human biopotential energy. Capacitive sensors provide accurate and stable measurements without requiring direct skin contact, conductive liquids or gels, or fixation, distinguishing them from other approaches [18]. They are also an excellent option for daily use because they are not sensitive to environmental factors [3].

4.1.2. Piezoresistive

These sensors consist of a conductive polymer that exhibits variations in electric resistance when subjected to external stress. More precisely, when the pressure increases, the resistance reduces [5]. Among the most used sensors in wearable technology is the FSR. There is an increasing fascination with utilizing FSRs that are integrated into shoes for the purpose of analyzing one’s walking pattern. This is mainly because FSRs are easy to wear and carry, making it convenient to gather data during regular daily tasks [26]. Nevertheless, several researchers contend that FSRs are not ideal for everyday practical use because of their restricted longevity and susceptibility to how they are affixed to the insole [15].

A recent study has documented the creation of a flexible textile piezoresistive sensor (TPRS) that has shown exceptional sensing capabilities. TPRSs demonstrate enhanced sensitivity in lower pressure ranges, a wider range of force detection, quicker response times, superior durability, and dependable detection across different pressure levels [14].

4.1.3. Piezoelectric

Piezoelectricity is the phenomenon where electric charge builds up on the surfaces of materials that lack a center of symmetry when they are subjected to mechanical force. Insole piezoelectric sensors typically comprise sandwich structures of electrodes and a piezoelectric film [5]. These sensors possess benefits such as minimal energy usage, uncomplicated structure, and exceptional sensitivity [18]. Nevertheless, these devices cannot detect static forces caused by the leakage current flowing through the amplifying circuit that follows, hence limiting their use in commercial products.

Piezoelectric pressure sensors can be classified into three categories according to the materials utilized: inorganic, organic, and composite. Combining piezoelectric materials with transistors makes it possible to reduce the interference caused by noise along the transmission line and improve the functionality of the piezoelectric pressure sensor. Piezoelectric film-based technology offers a practical option for measuring shear stresses or forces in many directions on the insole of the foot [5].

4.2. Inertial Measurement Unit(s)

Inertial measurement units are the most efficient techniques for measuring gait phases and events. These devices possess the qualities of portability, energy efficiency, cost-effectiveness, durability, and reliability [15]. IMUs combine accelerometers and gyroscopes to detect linear acceleration and angular velocity [33]. In addition to accelerometers and gyroscopes, certain IMUs also include magnetometers. These devices are inertial-magnetic or magneto-inertial measurement (MIMU) [1]. Driven by recent advancements in embedded technology within human–machine systems, IMU sensors are now embedded in smart and wearable devices such as shoes, watches, phones, and adhesive bandages to support daily human activities [17].

Accurate and dependable real-time identification of gait events using IMUs are crucial for the development of clinically significant gait characteristics that can differentiate between normal and impaired gait. Precision is also essential for developing individualized gait rehabilitation techniques for patients and controlling prosthetic devices based on feedback from gait phases [17]. Achieving precise human motion localization using IMU sensors is difficult due to various issues, including sensor drift, the sensor’s orientation with respect to the Earth, and magnetic interference from the surrounding environment [17].

4.3. Other Types of Sensors

Following the discussion on the various types of sensors and their prevalent use, it is crucial to highlight that the literature contains evidence and studies on a diverse range of sensors. For example, a novel conductive bamboo fabric was developed and can be conveniently adapted into various high-sensitivity, low-response-time wearable sensors. Applications permit the utilization of finger gloves to sense and control remote manipulation fingers, as well as waterproof sensing insoles to record human walking gestures. The study of the connections between fabric resistance and force response establishes a basis for developing and personalizing wearable sensors utilizing the conductive fabric. This technology has great promise for use in digital devices, physical activities, and healthcare applications [19].

Additionally, there are reports that a fully functional 3D active foot insole was developed and tested to assess the pressure-sensing capabilities. The results confirm that the proposed sensor possesses considerable potential for real-world applications such as rehabilitation, wearable devices, soft robotics, smart clothing, gait analysis, motion capture, augmented reality (AR)/virtual reality (VR), and numerous other fields [20].

4.4. Combination of Sensors and Systems

A novel addition to the market of wearable sensors is the integration of an insole pressure sensor with an IMU. This configuration exploits the advantages of both types of sensors [33]. While the combination of pressure and inertial sensors can improve performance, it also comes with drawbacks in terms of design layout, potential redundancy, higher power consumption, and reduced robustness [3].

However, other studies emphasize the benefits of incorporating different sensors and systems. Thi Thu Vu et al. [15] found that when IMU are combined with pressure insoles, the accuracy of gait phase detection algorithms improves. This allows the identification of additional sub-phases during stance and swing and provides measurements for pressure, angular velocity, and linear acceleration. A separate study demonstrated that integrating insole and tendon sensors [24] yielded more precise forecasts for walking and calf raises activities than using separate sensors [24]. In addition, a collection of piezoelectric pressure sensors made from electronic textiles was incorporated into the insoles to quantify plantar pressure. An IMU, consisting of a three-axis accelerometer, three-axis gyroscope, and three-axis magnetometer, was used to record gait characteristics while in motion [29].

4.5. Sensor Placement

Before sensor placement, a sensitivity analysis is essential to evaluate the performance of gait event detection and determine whether it depends on the IMU’s location on the foot [15]. Sensor misplacement can be related to two types of errors. The first type involves the sensor being placed on the correct segment but with an arbitrary orientation relative to the underlying bone. The second type, which is the focus of this study, occurs when the sensor is placed on the incorrect segment, such as switching between the left and right limbs or between the lower and upper limbs.

Although a verification procedure could mitigate this error, it imposes an additional burden on the user, who may lack technical expertise. Furthermore, it extends the installation time, which should be minimized, particularly for patient measurements. Human error is also a concern, especially in multi-sensor applications where each sensor must be attached to a specific body segment. This error becomes even more significant in scenarios where patients or their caregivers are responsible for sensor installation [22].

Gyroscope orientation remains constant relative to the body segment, and gyroscopes are unaffected by gravity and less prone to noise. In contrast, accelerometers are noisier, influenced by gravity, and sensitive to both position and orientation [33].

Unlike IMU, sensor placement is not a significant issue for insole pressure sensors. While IMU can be placed anywhere on the body, insole pressure sensors are typically located within the subject’s shoe. The conventional method for placing FSR in an insole involves positioning them at specific pressure points, such as the heel, toe, and first and fifth metatarsals. These insole pressure sensors require the correct foot size to ensure proper alignment with the pressure points [33].

Some products in the wearable sensor market have fixed the IMU position relative to the insole pressure sensor, thereby reducing errors caused by variations in IMU placement between and within segments and across different datasets and subjects [33].

A conventional approach involves positioning the sensors at predefined locations. This method includes three steps: First, the plantar surface is segmented into several regions based on foot structure or function. Second, specific regions are selected according to the characteristics of the target pathology. Third, within these chosen regions, the positions most associated with disease progression, such as peak stress points or symmetrical positions on the medial and lateral aspects of the foot, are identified [5].

4.6. Sensor Size and Number

The primary design consideration is the sensor size, which is directly related to the quality of raw data collected [5]. According to Chen et al. [5], an insole sensor should possess a compact size, lightweight construction, and high flexibility. Additionally, it must exhibit an appropriate force-sensing range and reliable responsiveness under extreme conditions. The material and design of the device should be suitable for its intended application area, and the system’s scanning speed must meet the application’s specific requirements.

A larger sensor may underestimate peak pressure values, while smaller sensors can make it challenging to track the displacement of points of interest during gait accurately. Consequently, it is recommended to use smaller sensors in an array configuration [16].

As for the number of sensors, a simplified design with the minimum number of integrated sensors is preferred, as cost, power consumption, and robustness are the primary considerations [3]. However, as the resolution of insole pressure sensors improves, the approach is transitioning towards integrating a higher density of sensors within the insole. This allows comprehensive data collection across the entire foot, enabling the identification of pressure hotspots through software analysis during signal processing [33]. Other authors argue that increasing the number of sensors allows the collection of a greater volume of information, thereby enhancing the accuracy of gait assessment. Besides sensor quantity, other factors influencing the accuracy of assessment outcomes include the precision of gait measurement parameters and the amount of data collected [25]. Unfortunately, the increase in the sensor number is reflected in the energy consumption and in the computational cost to analyze and process the signals to converge to a solution [16].

The plantar surface can be segmented into distinct regions to gather more pertinent data for pathology analysis [16]. Next, as an example, we can observe in Figure 8, the segmentation, the divided zones and the position of each sensor in an insole.

Thus, the minimum number of recommended sensors is fifteen. Still, this number can vary according to foot size, leading to the displacement of key pressure points [34] For example, in Tan et al.’s [14] work, six piezoresistive sensors were installed beneath the primary weight-bearing areas of the insole, specifically the big toe, the first and fourth metatarsal heads, the midfoot, and the heel.

4.7. Applications

These devices aim to assist researchers and medical professionals by generating reports that facilitate the study of patient progress and aid in identifying pathologies. Additionally, they are valuable for designing and testing orthotics, enhancing athletic techniques, and providing information about treatments, thereby promoting further research. Numerous reported developments have been conducted with future applications in mind, including monitoring, pattern extraction, rehabilitation, and disorder detection, among others [16].

Intelligent insoles have effectively monitored gait cycles and/or plantar pressure [31]. Gait is a rhythmic and periodic movement necessitating coordination, balance, and synchronization, facilitated by the proper functioning of the central and peripheral nervous systems. Movement is driven by ground reaction forces (GRF) applied to the body through its contact with the ground. Just as individuals have unique fingerprints, they also possess distinctive gait characteristics and movement patterns. Aging impacts the dynamics of foot pressure distribution during normal walking. In elderly individuals, the forces and pressures exerted on the medial foot regions are diminished, leading to reduced propulsion during the transition from the heel strike phase to the toe-off phase [26].

The analysis of human movement encompasses the concepts of kinetics and kinematics. Kinetics describes the causes of movement, focusing on the forces (mass and acceleration), torques, and power generated. In contrast, kinematics focuses on the linear and angular descriptions of movement, including changes in velocity, position, displacement, and acceleration over time. Consequently, gait analysis requires a quantitative study of force parameters and temporal and spatial parameters by calculating key characteristics. Recognizing the importance of gait analysis is essential for detecting and managing various neurological and musculoskeletal disorders. Identifying gait events is valuable for enhancing gait analysis, developing precise monitoring systems, and evaluating treatments for pathological gait [10].

In clinical practice, the CoP is frequently utilized to assess an individual’s postural control stability, which is closely linked to various neurological diseases and movement disorders, such as Alzheimer’s disease, Parkinson’s disease, and chronic ankle instability. These conditions typically involve extended development or rehabilitation processes, necessitating long-term CoP monitoring [21].

4.7.1. Rehabilitation

In rehabilitation, electronic systems are highly effective for monitoring and analysis, as they exhibit greater sensitivity to changes in patient improvement than traditional medical observation. Insole sensor systems designed for these applications are especially user-friendly since they do not require controlled environments and are often cost-effective for implementation outside clinical settings. Moreover, these compact and convenient sensors offer practical alternative to larger and more heavy equipment such as force plates, motion capture systems, and instrumented treadmills [16].

The main recovery application is the rehabilitation of the limb, especially in post-surgery phase. In postoperative recovery the wearables can provide comprehensive monitoring capabilities [1]. Real-time gait detection using wearable sensors offers an unprecedented approach for delivering clinical interventions to individuals with gait impairments (stiffness, balance issues, unsteady and staggering walking, shuffling steps, difficulty starting and stopping movement, and other behaviors). A normal gait pattern indicates repeated gait cycles, and an abnormal gait pattern may suggest a neuron disease, a lower limb motor disfunction, and injuries [3]. The optimal combination of sensors and gait detection methodologies enables the creation of assistive devices that can significantly enhance the effectiveness of rehabilitation and improve the quality of life for people with ambulatory deficits [33]. To this matter the development of algorithms have the ability and capacity to extract data of interest, such as CoP, contact time, and cadence, parameters that are commonly associated to development of diseases [5].

To demonstrate the importance of WT in rehabilitation, in a study [25], there was utilized seven IMU sensors and a pair of plantar pressure insoles to collect walking data from both healthy individuals and patients with hemiplegia. Adding to the previous study, the works of Chen et al. [5] and Almuteb et al. [3] present several examples of how important it is to gather and analyze various metrics such as pressure, joint angular velocity, stride interval, swing interval, stance interval of both feet, and acceleration of subjects with and without diseases to monitoring the status and conditions. Diseases such as Parkinson’s disease, Huntington’s disease, and amyotrophic lateral sclerosis is referred in the previous work. Yang et al. [1] also confirm that limb and movement disorders allow motion sensors to be used, such as postural instability (including falls) of patients with Parkinson’s disease. This kind of equipment can be helpful for diabetics by monitoring foot pressure, temperature, humidity alterations of foot ulcers in daily life [3].

In conclusion, we can say with some certainty that the more data, the better. However, in rehabilitation, there are many diseases and pathologies to consider; therefore, it is very difficult to collect data, and there is no basis for comparison to analyze and make assumptions. Chen et al. [5] have noted the following: “To extract the effect of a specific disease on patients’ gait from multiple diseases fusion, it is necessary to track the gait of different patients for a long time and continuously analyze the gait of the same patient.”

4.7.2. Performance

The use and assessment of biomechanical wearables in sports embody a wide range of athletic activities and movements. These devices are applicable to various actions, including running, jumping, throwing, speed, acceleration, balance, lower body symmetry, force, pressure, and impact [2]. The most common on-field application is sports activity recognition. The data acquired related to body posture and limb coordination serves as an important parameter of an athlete’s motion [1]. Additionally, the acquisition of temporal-spatial parameters of bipedal human locomotion is fundamental for several disciplines and are essential for balance, postural stability, and athletic performance evaluation [3]. In sports, human gait analysis is used to monitor and improve the performance of athletes [12]. Gait analysis can be performed with vision-based or sensor-based methods. Vision-based requires specialized laboratories and personnel and therefore limits the possible test conditions. In contrast, pressure insoles are less expensive, ensure high-flexibility, and offer the possibility of performing field measurement.

In sports performance, it is possible to rely on the data acquired and given by WT to track a player’s position on the field and carry out a tactical analysis or even estimate the external measurements of loads in specific movements to help prevent injuries [18].

It is important to state that the WT are useful not only for athletes but also for coaches and other personnel included in a club staff. As referred in Amaro’s study [4] about the distribution of plantar pressure on both feet, with the information about the exercises and movements that are more conductive and have a higher probability of injury occurrence, the coaching staff have the possibility of planning an exercise or a session with more attention.

5. Discussion

The integration of multiple sensors enhances data accuracy, reliability, and functionality, offering improved spatial and temporal resolution, redundancy, and comprehensive environmental monitoring capabilities across various scientific and industrial applications.

Results obtained with laboratory solutions are more accurate; on the other hand, wearable solutions offer the advantage of estimating temporal parameters during regular training sessions [12]. In Elstub, L. J. et al. [8] study, it was possible to estimate peak tibial bone force during running with an IMU system combined with a pressure-sensing insole and a trained algorithm with average errors of 5,7%. The authors conclude that it is necessary to improve and determine calibration methods to reduce the percentage of errors and to estimate the forces with more precision. In another study [24], the results obtained show that a low-cost pressure insole and tendon sensor can produce estimates similar to the reported accuracy of commercial devices.

Insole data offers high precision and accuracy in estimations, facilitating detailed analysis of gait patterns, pressure distribution, and biomechanical metrics, which is crucial for advanced clinical assessments and optimizing sports performance. For the ankle joint torque estimations, we found that the insole data was generally more accurate than the ankle angle from inverse kinematics or the tendon sensor. Additionally, the insole data alone were accurate for walking. However, it was insufficient for the calf to raise task [24].

The validation of market-available equipment ensures accuracy, reliability, and compliance with industry standards, enhancing performance consistency and user safety, which is crucial for scientific research, medical applications, and industrial processes. For example, a study showed the validation of the gait parameters from FeetMe insoles under single and cognitive dual-task conditions, with promising results for future applications. Their flexibility, portability, and ease of use make them relevant tools for large-scale analyses of patients with Parkinson’s disease [9].

In this study, it is important to understand what has been accomplished so far, define what is on the right track and what urgently needs to be improved, and do so in future works and research. Nevertheless, there has been no conclusion on which sensor type or position is the best for a specific application. The data of most studies lack diversity. Moreover, data collection is mostly carried out in fully controlled environments. Therefore, the data collected and analyzed by the studies may not represent daily life conditions due to the limited amount of data. Multi-discipline research should be encouraged to overcome this limitation, connecting engineers to doctors and healthcare workers to collect more data for analysis and improve robustness. From the perspective of applications, gait studies are the main focus. A diverse subject database, including sensor readings from various types and positions, should be collected to compare the prototypes and evaluate the performance with the same data [3].

Regarding hardware, Chen et al.’s work [26] identified the necessity, in future work, of optimizing sensor technology to reduce the error in the estimation of the force and acceleration values to improve the wireless communication range [29], as well as the need to develop dynamic sensor-driven power management techniques to extend battery life [29]. Other works [1,17] pointed out the necessity of improving the ergonomics; size; manufacturing and cost; and battery technology, optimization and storage of wearable devices. The use and necessity of equipping the insole with GPS is also referred in Lin et al. [29] work, especially in adults with Alzheimer’s disease, dementia, autism, or other cognitive disorders and people who need real-time location information. The authors also mention that when conducting a study not limited to a gait laboratory, it is essential to experiment and study the equipment in real life, such as indoor walking, outdoor exercises, climbing hills, ascending/descending stairs, etc.

The provision of real-time data offers substantial advantages in performance analysis, injury prediction, and rehabilitation. Harnessing these technologies to bolster athletes’ health, recovery, and performance can substantially expand access to advanced training and preparation techniques within the sports industry. Nevertheless, it is imperative to address sensitive issues such as equity, privacy, security, and intellectual property rights in utilizing these technologies [18].

Insole gait monitoring systems have not yet been successfully applied in daily life. Three reasons can be appointed to this reason: (1) raw data characteristics; (2) energy consumption; and (3) requirement of large datasets for disease emulation. One big problem that remains till these days is the lack of offer in the current market of a low-cost design [5].

6. Research Challenges and Future Perspectives

Despite all the notable advancements in smart insole technology, there are still some challenges that must be overcome to further progress in this field.

A significant limitation is the requirement for thorough validation of smart insole technology according to recognized laboratory standards to guarantee its reliability and accuracy. It is essential to develop standardized testing techniques that can be broadly implemented. For that purpose, it is important to perform comparative research between smart insoles and conventional gait analysis equipment such as force plates and motion capture devices. Finally, submitting validation studies to peer-reviewed journals will enhance the credibility of the research.

The broad adoption of advanced smart insoles may be limited by the high costs associated with them. To overcome this issue, it is necessary to investigate cost-cutting methods such as implementing mass production techniques and utilizing more affordable materials while maintaining high standards of quality. Performing cost–benefit studies to emphasize the enduring cost savings and advantages of utilizing intelligent insoles for preventative healthcare and rehabilitation can demonstrate how valuable they are. To reach this goal, it is crucial to explore financial opportunities or incentives for the creation of new prototypes and equipment, particularly in the field of healthcare.

The lack of variety in the data limits the capacity to apply research findings to a wider context. Developing a centralized repository to store and distribute raw data from different research projects, with a focus on integrating a wide range of populations and situations, will be very important. It is fundamental to encourage multi-center research that will gather data from various demographic and clinical contexts and to establish standardized techniques for collecting data. These measures will allow the consolidation of data, enable comparative studies, and allow progress in the various scientific fields.

One of the most important objectives is to prioritize privacy and ensure data security while also guaranteeing that technology remains easily accessible and user-friendly. Implementing strong encryption and confidentiality methods is vital for protecting user data. Defining clear and explicit policies and processes addressing data ownership, sharing, and consent is fundamental.

Optimizing comfort and efficiency demand improvements in personalized wearable technology devices. Providing resources to research focused on developing customized insoles tailored to the distinct anatomical and functional needs of each person is extremely significant. The use of 3D printing and adaptive materials can improve the user experience by creating customized insole designs. It is also necessary to carry out user-centered design studies to constantly improve comfort and functionality, considering user feedback as a premise.

The integration of advanced data analytics techniques remains insufficient. As a way to deal with this limitation, it is necessary to develop and implement machine learning techniques that can enhance the precision and predictive capabilities of smart insoles. Using artificial intelligence to analyze extensive datasets with the objective of identifying patterns, detecting anomalies, and constructing predictive models will result in deeper insights. Collaborating with data scientists to incorporate advanced analytics into insole systems can be beneficial for obtaining real-time feedback and monitoring.

Another problem emerges from the insufficient integration and compatibility with other wearable devices. It is imperative to comprehend the existing options and establish operational procedures to enable effective data transmission between smart insoles and other wearable devices such as smartwatches and fitness trackers. Developing integrated platforms that can gather data from different sources to deliver comprehensive health and performance studies will improve the effective use of these technologies. Establishing collaborations with companies specialized in diverse wearable technologies could facilitate the development of integrated health monitoring systems.

The shortage of real-world validation and longitudinal investigations is a significant limitation. It is essential to design and carry out long-term studies that monitor the performance and effects of smart insoles in real-life environments. It is necessary to incorporate a wide range of contexts, including indoor and outdoor settings as well as different surfaces, in order to determine the technology’s robustness and versatility. Conducting long-term monitoring of users to evaluate the strength and long-lasting advantages of smart insoles could provide significant insights.

By focusing on these particular areas, future research has the potential to significantly progress the field of smart insole technology, improving both its scientific basis and its practical uses.

7. Conclusions

The thorough examination of WT in sports and rehabilitation highlights a notable advancement in the use of biomechanics to improve human performance and enhance risk analysis procedures. Exoskeletons and other sensors, such as IMUs and pressure sensors, are essential in both industries, as they gather vital data for improving performance, monitoring diseases or disorders, and evaluating risks. The sports industry utilizes WT to a greater extent in various activities, such as running, leaping, and other athletic activities. The primary focus is on improving individual performance metrics, specifically speed, balance, and force. On the other hand, monitoring the status of a disease such as diabetes or analyzing the walking pattern of a person with a physical disorder could be among the applications in the rehabilitation or health field.

Both the sports and rehabilitation sectors are increasingly adopting WT solutions, which offer mutual benefits through shared improvements in data analysis and implementation methodologies. Although there have been significant breakthroughs, there also exist obstacles that still need to be addressed. These challenges include the requirement for validation against laboratory standards and worries about the cost, which could potentially obstruct the broad use of more accurate wearable devices. For that reason, it is crucial to augment the available data with information about the various metrics that are possible to collect and analyze, involving different kinds of subjects. Furthermore, concerns regarding privacy, data security, and fair access to technology are of utmost importance and require strong protection measures to guarantee the safe and efficient implementation of WT. Further investigation is warranted to delve into the advancement of customized WT devices that optimize both comfort and performance and to scrutinize the expanding realm of wearable imaging technology. These developments have the potential to transform sports rehabilitation and performance monitoring completely. They provide new knowledge and abilities that contribute to the health and recovery of athletes. Continuous evaluations of cutting-edge technologies and their practical uses will be crucial for health and safety decision makers to maximize the benefits of innovative tools for improving athlete performance and safety in the sports industry. Another important factor can be the implementation of machine learning techniques to improve the precision of the estimations acquired, compensate the limitations of the devices utilized, and explore their respective specifications.

Our initial objective was to answer the following question: “What types of sensors are employed in insoles for the purposes of enhancing performance and improving rehabilitation?” In this study, we concluded that most of the studies or the equipment utilized were pressure sensors and IMUs. However, is certain that there are still more tests and studies to make with various types of sensors and in different situations. It is also determined that there is not an agreement in the scientific community about the quantity of sensors and the position of each sensor.