Wearable Inertial Sensors Research Articles

Analysis of the Recent AI for Pedestrian Navigation With Wearable Inertial Sensors

Wearable devices embedding inertial sensors enable autonomous, seamless, and low-cost pedestrian navigation. Appealing it is, the approach faces several challenges: measurement noises, different device-carrying modes, different user dynamics, and individual walking characteristics. Recent research applies Artificial Intelligence (AI) to improve inertial navigation's robustness and accuracy. Our analysis identifies 2 main categories of AI approaches depending on the inertial signals segmentation: either using human gait events (steps or strides) or fixed-length inertial data segments. A theoretical analysis of the fundamental assumptions is carried out for each category. Two state-of-the-art AI algorithms (SELDA, RoNIN), representative of each category, and a gait-driven non-AI method (SmartWalk) are evaluated in a 2,17 km long open access dataset, representative of the diversity of pedestrians' mobility surroundings (open-sky, indoors, forest, urban, parking lot). SELDA is an AI-based stride length estimation algorithm, RoNIN is an AI-based positioning method, and SmartWalk is a gait-driven non-AI positioning method. The experimental assessment shows the distinct features in each category and their limits with respect to the underlying hypotheses. On average, SELDA, RoNIN, and SmartWalk achieve 8 m, 22 m, and 17 m average positioning errors (RMSE) respectively, on six testing tracks recorded with two volunteers in various environments.

Automatic Recognition of Construction Worker Activities Using Deep Learning Approaches and Wearable Inertial Sensors

The automated evaluation and analysis of employee behavior in an Industry 4.0-compliant manufacturing firm are vital for the rapid and accurate diagnosis of work performance, particularly during the training of a new worker. Various techniques for identifying and detecting worker performance in industrial applications are based on computer vision techniques. Despite widespread computer vision-based approaches, it is challenging to develop technologies that assist the automated monitoring of worker actions at external working sites where camera deployment is problematic. Through the use of wearable inertial sensors, we propose a deep learning method for automatically recognizing the activities of construction workers. The suggested method incorporates a convolutional neural network, residual connection blocks, and multi-branch aggregate transformation modules for high-performance recognition of complicated activities such as construction worker tasks. The proposed approach has been evaluated using standard performance measures, such as precision, F1-score, and AUC, using a publicly available benchmark dataset known as VTT-ConIoT, which contains genuine construction work activities. In addition, standard deep learning models (CNNs, RNNs, and hybrid models) were developed in different empirical circumstances to compare them to the proposed model. With an average accuracy of 99.71% and an average F1-score of 99.71%, the experimental findings revealed that the suggested model could accurately recognize the actions of construction workers. Furthermore, we examined the impact of window size and sensor position on the identification efficiency of the proposed method.

Classification and comparison of human activities by machine learning

The current study of human activity recognition and classification has been an important part of promoting the development of science and technology in society. Human activity recognition and classification are in several fields, such as competitive sports, criminal investigation field, etc. As the field of micro-electromechanics continues to evolve, more accurate human recognition is becoming possible, with wearable multi-axis inertial sensors that allow us to visually detect the desired data. In this paper, the data of 19 human activities for 8 testers are feature extracted and normalized. The data are divided into training and test sets by machine learning models: support vector machine (SVR) classification, XGBoost classification, and logistic regression. The experiment was repeated 10 times to take the average value. The models were then scored, and by comparing integrated machine learning with traditional machine learning, it was found that integrated learning improved by 5%−29% in terms of accuracy compared to traditional machine learning.

Effect of taVNS on turning in advanced Parkinson's disease patients – a wearable inertial motion sensor study

Effect of taVNS on turning in advanced Parkinson's disease patients – a wearable inertial motion sensor study

Longitudinal Evaluation of Hemiplegic Ankle Rehabilitation Efficacy by Wearable Inertial Sensor Systems with an Assortment of Machine Learning Algorithms

Longitudinal Evaluation of Hemiplegic Ankle Rehabilitation Efficacy by Wearable Inertial Sensor Systems with an Assortment of Machine Learning Algorithms

Pre-Impact and Impact Fall Detection Based on a Multimodal Sensor Using a Deep Residual Network

Falls are the contributing factor to both fatal and nonfatal injuries in the elderly. Therefore, pre-impact fall detection, which identifies a fall before the body collides with the floor, would be essential. Recently, researchers have turned their attention from post-impact fall detection to pre-impact fall detection. Pre-impact fall detection solutions typically use either a threshold-based or machine learning-based approach, although the threshold value would be difficult to accurately determine in threshold-based methods. Moreover, while additional features could sometimes assist in categorizing falls and non-falls more precisely, the estimated determination of the significant features would be too time-intensive, thus using a significant portion of the algorithm’s operating time. In this work, we developed a deep residual network with aggregation transformation called FDSNeXt for a pre-impact fall detection approach employing wearable inertial sensors. The proposed network was introduced to address the limitations of feature extraction, threshold definition, and algorithm complexity. After training on a large-scale motion dataset, the KFall dataset, and straightforward evaluation with standard metrics, the proposed approach identified pre-impact and impact falls with high accuracy of 91.87 and 92.52%, respectively. In addition, we have investigated fall detection’s performances of three state-of-the-art deep learning models such as a convolutional neural network (CNN), a long short-term memory neural network (LSTM), and a hybrid model (CNN-LSTM). The experimental results showed that the proposed FDSNeXt model outperformed these deep learning models (CNN, LSTM, and CNN-LSTM) with significant improvements.

A portable and reliable tool for on-site physical reaction time (RT) measurement

The drop-stick test system (DTS) invention has the capability to measure reaction accurately for sideline mild traumatic brain injury (mTBI) assessment. The reaction time (RT) measurements showed moderate to good inter-instrument reliability with an overall ICC of 0.82 (95 % CI 0.78–0.85). RT is a useful biomarker of mTBI or concussion, but existing technologies in controlled laboratory environments are not feasible for assessments in the field. With wearable technologies and wireless connection with smartphones, it is now easier to conduct RT assessments on the field. The purpose was to develop a portable DTS involving wearable inertial sensors translatable from the laboratory to the sideline assessments on athletic fields. This project investigated if RT measurements using sensors are reliable metrics when compared to the gold-standard motion capture laboratory-test. The intent was to invent and assess the reliability of a clinical DTS using sensors in conjunction with an easy RT measurement protocol. The integrated inertial sensors in the DTS have the potential to reliably capture RT, allowing it to more easily translate into the training room or field environment.

Spike encoding techniques for IoT time-varying signals benchmarked on a neuromorphic classification task.

Spiking Neural Networks (SNNs), known for their potential to enable low energy consumption and computational cost, can bring significant advantages to the realm of embedded machine learning for edge applications. However, input coming from standard digital sensors must be encoded into spike trains before it can be elaborated with neuromorphic computing technologies. We present here a detailed comparison of available spike encoding techniques for the translation of time-varying signals into the event-based signal domain, tested on two different datasets both acquired through commercially available digital devices: the Free Spoken Digit dataset (FSD), consisting of 8-kHz audio files, and the WISDM dataset, composed of 20-Hz recordings of human activity through mobile and wearable inertial sensors. We propose a complete pipeline to benchmark these encoding techniques by performing time-dependent signal classification through a Spiking Convolutional Neural Network (sCNN), including a signal preprocessing step consisting of a bank of filters inspired by the human cochlea, feature extraction by production of a sonogram, transfer learning via an equivalent ANN, and model compression schemes aimed at resource optimization. The resulting performance comparison and analysis provides a powerful practical tool, empowering developers to select the most suitable coding method based on the type of data and the desired processing algorithms, and further expands the applicability of neuromorphic computational paradigms to embedded sensor systems widely employed in the IoT and industrial domains.

Acceptability of wearable inertial sensors, completeness of data, and day-to-day variability of everyday life motor activities in children and adolescents with neuromotor impairments.

Monitoring the patients' motor activities in a real-world setting would provide essential information on their functioning in daily life. In this study, we used wearable inertial sensors to monitor motor activities of children and adolescents with congenital and acquired brain injuries. We derived a set of clinically meaningful performance measures and addressed the following research questions: Is the target population willing to wear the sensors in their habitual environment? Which factors lead to missing data, and can we avoid them? How many measurement days are needed to obtain reliable estimates of the children's and adolescents' motor performance? The study participants wore our sensor system for seven consecutive days during waking hours. First, we derived the daily hand use of all participants, the duration of different body positions and the wheeling activity of individuals using a manual wheelchair, and walking-related measures in individuals being able to walk. Then, we analyzed the reasons for missing data and determined the reliability of the performance measures mentioned above. The large majority (41 of 43 participants) was willing to wear the sensor system for a week. However, forgetting to reattach the sensors after charging them overnight and taking them off during bathing and swimming was the main contributor to missing data. Consequently, improved battery life and waterproofness of the sensor technology are essential requirements for measurements in daily life. Besides, 5 of 11 performance measures showed significant differences between weekdays and weekend days. The reliability, measured with the intraclass correlation coefficient, ranged between 0.82 and 0.98. Seven measurement days were enough to obtain significantly higher reliability scores than the desired level of 0.8 for all but two performance measures. In children and adolescents with neuromotor impairments, we recommend monitoring everyday life motor activities on seven consecutive days. The target population accepted this measurement protocol, it covers school days and weekend days, and the number of measurement days is sufficient to obtain reliable estimates of motor performance.

Using Wearable Inertial Sensors to Assess Mobility of Patients With Hematologic Cancer and Associations With Chemotherapy-Related Symptoms Before Autologous Hematopoietic Stem Cell Transplant: Cross-sectional Study.

Wearable sensors could be a simple way to quantify and characterize mobility in patients with hematologic cancer scheduled to receive autologous hematopoietic stem cell transplant (autoHSCT) and how they may be related to common treatment-related symptoms and side effects of induction chemotherapy. We aimed to conduct a cross-sectional study comparing mobility in patients scheduled to receive autoHSCT with that in healthy, age-matched adult controls and determine the relationships between patient mobility and chemotherapy-related symptoms. Patients scheduled to receive autoHSCT (78/156, 50%) and controls (78/156, 50%) completed the prescribed performance tests using wearable inertial sensors to quantify mobility including turning (turn duration and number of steps), gait (gait speed, stride time, stride time variability, double support time, coronal trunk range of motion, heel strike angle, and distance traveled), and balance (coronal sway, coronal range, coronal velocity, coronal centroidal frequency, sagittal sway, sagittal range, sagittal velocity, and sagittal centroidal frequency). Patients completed the validated patient-reported questionnaires to assess symptoms common to chemotherapy: chemotherapy-induced peripheral neuropathy (Functional Assessment of Cancer Therapy/Gynecologic Oncology Group-Neurotoxicity subscale), nausea and pain (European Organization for Research and Treatment of Cancer Quality of Life Questionnaire), fatigue (Patient-Reported Outcomes Measurement Information System Fatigue Short Form 8a), vertigo (Vertigo Symptom Scale-short form), and depression (Center for Epidemiological Studies-Depression). Paired, 2-sided t tests were used to compare mobility between patients and controls. Stepwise multivariable linear regression models were used to evaluate associations between patient mobility and symptoms. Patients aged 60.3 (SD 10.3) years had significantly worse turning (turn duration; P<.001), gait (gait speed, stride time, stride time variability, double support time, heel strike angle, stride length, and distance traveled; all P<.001), and balance (coronal sway; P<.001, range; P<.001, velocity; P=.02, and frequency; P=.02; and sagittal range; P=.008) than controls. In patients, high nausea was associated with worse stride time variability (ß=.001; P=.005) and heel strike angle (ß=-.088; P=.02). Pain was associated with worse gait speed (ß=-.003; P=.003), stride time variability (ß=.012; P=.02), stride length (ß=-.002; P=.004), and distance traveled (ß=-.786; P=.005). Nausea and pain explained 17% to 33% and 14% to 36% of gait variance measured in patients, respectively. Patients scheduled to receive autoHSCT demonstrated worse mobility in multiple turning, gait, and balance domains compared with controls, potentially related in part to nausea and pain. Wearable inertial sensors used in the clinic setting could provide granular information about mobility before further treatment, which may in turn benefit from rehabilitation or symptom management. Future longitudinal studies are needed to better understand temporal changes in mobility and symptoms across the treatment trajectory to optimally time, design, and implement strategies, to preserve functioning in patients with hematologic cancer in the long term.

Assessing Cerebellar Disorders with Wearable Inertial Sensor Data Using Time-Frequency and Autoregressive Hidden Markov Model Approaches.

Wearable sensor data is relatively easily collected and provides direct measurements of movement that can be used to develop useful behavioral biomarkers. Sensitive and specific behavioral biomarkers for neurodegenerative diseases are critical to supporting early detection, drug development efforts, and targeted treatments. In this paper, we use autoregressive hidden Markov models and a time-frequency approach to create meaningful quantitative descriptions of behavioral characteristics of cerebellar ataxias from wearable inertial sensor data gathered during movement. We create a flexible and descriptive set of features derived from accelerometer and gyroscope data collected from wearable sensors worn while participants perform clinical assessment tasks, and use these data to estimate disease status and severity. A short period of data collection (<5 min) yields enough information to effectively separate patients with ataxia from healthy controls with very high accuracy, to separate ataxia from other neurodegenerative diseases such as Parkinson’s disease, and to provide estimates of disease severity.

A Narrative Review on Wearable Inertial Sensors for Human Motion Tracking in Industrial Scenarios

Industry 4.0 has promoted the concept of automation, supporting workers with robots while maintaining their central role in the factory. To guarantee the safety of operators and improve the effectiveness of the human-robot interaction, it is important to detect the movements of the workers. Wearable inertial sensors represent a suitable technology to pursue this goal because of their portability, low cost, and minimal invasiveness. The aim of this narrative review was to analyze the state-of-the-art literature exploiting inertial sensors to track the human motion in different industrial scenarios. The Scopus database was queried, and 54 articles were selected. Some important aspects were identified: (i) number of publications per year; (ii) aim of the studies; (iii) body district involved in the motion tracking; (iv) number of adopted inertial sensors; (v) presence/absence of a technology combined to the inertial sensors; (vi) a real-time analysis; (vii) the inclusion/exclusion of the magnetometer in the sensor fusion process. Moreover, an analysis and a discussion of these aspects was also developed.

Validity of the Baiobit Inertial Measurements Unit for the Assessment of Vertical Double- and Single-Leg Countermovement Jumps in Athletes.

Jump tests are simple, quick to execute, and considered the most reliable tool to measure lower extremities power and explosiveness in athletes. Wearable inertial sensors allow the assessment of jumping performance on any surface. The validity of inertial sensors measurements is a pivotal prerequisite to reliably implement their utilization in the clinical practice. Twenty-seven athletes (20 M/7 F, age: 27 ± 7 years old) performed five double-leg countermovement jumps (CMJs) and three single-leg CMJs per side with their hands on their hips. Jump height was measured/computed simultaneously with the optoelectronic system, force platforms, and the Baiobit inertial sensor system. The athletes completed the international physical activity questionnaire (IPAQ). When comparing the methods (Baiobit vs. force platforms), a non-statistically significant bias of 1.8 cm was found for two-leg CMJs and -0.6 cm for single-leg CMJs. The intraclass correlation coefficients (ICCs) was "excellent" for double-leg CMJs (ICC = 0.92, 95% CI = 0.89-0.94) and "good" for single-leg CMJs (ICC = 0.89, 95% CI = 0.85-0.91). When comparing the methods (Baiobit vs. force platforms + optoelectronic system), a non-statistically significant bias of -0.9 cm was found for two-leg CMJs and -1.2 cm for single-leg CMJs. The intraclass correlation coefficient (ICC) was "good" for both double-leg CMJs (ICC = 0.80, 95% CI = 0.73-0.85) and for single-leg CMJs (ICC = 0.86, 95% CI = 0.80-0.89). Baiobit tends to overestimate double- and single-leg CMJ height measurements; however, it can be recommended in the world of rehabilitation and sport analysis.

A Novel Deep Neural Network Method for HAR-Based Team Training Using Body-Worn Inertial Sensors.

Human activity recognition (HAR) became a challenging issue in recent years. In this paper, we propose a novel approach to tackle indistinguishable activity recognition based on human wearable sensors. Generally speaking, vision-based solutions struggle with low illumination environments and partial occlusion problems. In contrast, wearable inertial sensors can tackle this problem and avoid revealing personal privacy. We address the issue by building a multistage deep neural network framework that interprets accelerometer, gyroscope, and magnetometer data that provide useful information of human activities. Initially, the stage of variational autoencoders (VAE) can extract the crucial information from raw data of inertial measurement units (IMUs). Furthermore, the stage of generative adversarial networks (GANs) can generate more realistic human activities. Finally, the transfer learning method is applied to enhance the performance of the target domain, which builds a robust and effective model to recognize human activities.

OVERSAMPLING TECHNIQUE-BASED DATA AUGMENTATION AND 1D-CNN AND BIDIRECTIONAL GRU ENSEMBLE MODEL FOR HUMAN ACTIVITY RECOGNITION

The success of the human activity recognition (HAR) algorithm of a wearable inertial sensor depends on the extraction of good temporal dynamics from the sequential motor movement of humans. In the past, a heuristic method was used for the feature extraction process. Recently, deep neural network (DNN)-based studies, in which this process is automatically performed in the model, have shown good results. The performance of the DNN models is good and almost all processes are automated; however, they have some limitations—data-imbalance problems in the dataset and the small data decrease the performance. In this study, the HAR public dataset was improved by introducing an oversampling technique-based data-augmentation algorithm, and a stacking ensemble model with a one-dimensional convolutional neural network and a bidirectional gated recurrent unit as the head was used for classification. The HAR public dataset used Wireless sensor data mining (WISDM) and the University of California, Irvine human activity recognition using smartphones data set (UCI-HAR). Consequently, the proposed model showed better performance than that of previous studies for both datasets. The accuracy of the proposed model was 99.49% and 99.48%; the F1-score of the model was 99.84% and 99.78%, respectively.

A novel approach for modelling and classifying sit-to-stand kinematics using inertial sensors.

Sit-to-stand transitions are an important part of activities of daily living and play a key role in functional mobility in humans. The sit-to-stand movement is often affected in older adults due to frailty and in patients with motor impairments such as Parkinson's disease leading to falls. Studying kinematics of sit-to-stand transitions can provide insight in assessment, monitoring and developing rehabilitation strategies for the affected populations. We propose a three-segment body model for estimating sit-to-stand kinematics using only two wearable inertial sensors, placed on the shank and back. Reducing the number of sensors to two instead of one per body segment facilitates monitoring and classifying movements over extended periods, making it more comfortable to wear while reducing the power requirements of sensors. We applied this model on 10 younger healthy adults (YH), 12 older healthy adults (OH) and 12 people with Parkinson's disease (PwP). We have achieved this by incorporating unique sit-to-stand classification technique using unsupervised learning in the model based reconstruction of angular kinematics using extended Kalman filter. Our proposed model showed that it was possible to successfully estimate thigh kinematics despite not measuring the thigh motion with inertial sensor. We classified sit-to-stand transitions, sitting and standing states with the accuracies of 98.67%, 94.20% and 91.41% for YH, OH and PwP respectively. We have proposed a novel integrated approach of modelling and classification for estimating the body kinematics during sit-to-stand motion and successfully applied it on YH, OH and PwP groups.

Sensor-based outcomes to monitor everyday life motor activities of children and adolescents with neuromotor impairments: A survey with health professionals.

In combination with appropriate data processing algorithms, wearable inertial sensors enable the measurement of motor activities in children's and adolescents' habitual environments after rehabilitation. However, existing algorithms were predominantly designed for adult patients, and their outcomes might not be relevant for a pediatric population. In this study, we identified the needs of pediatric rehabilitation to create the basis for developing new algorithms that derive clinically relevant outcomes for children and adolescents with neuromotor impairments. We conducted an international survey with health professionals of pediatric neurorehabilitation centers, provided them a list of 34 outcome measures currently used in the literature, and asked them to rate the clinical relevance of these measures for a pediatric population. The survey was completed by 62 therapists, 16 doctors, and 9 nurses of 16 different pediatric neurorehabilitation centers from Switzerland, Germany, and Austria. They had an average work experience of 13 ± 10 years. The most relevant outcome measures were the duration of lying, sitting, and standing positions; the amount of active self-propulsion during wheeling periods; the hand use laterality; and the duration, distance, and speed of walking periods. The health profession, work experience, and workplace had a minimal impact on the priorities of health professionals. Eventually, we complemented the survey findings with the family priorities of a previous study to provide developers with the clinically most relevant outcomes to monitor everyday life motor activities of children and adolescents with neuromotor impairments.

Accuracy and comparison of sensor-based gait speed estimations under standardized and daily life conditions in children undergoing rehabilitation

BackgroundGait speed is a widely used outcome measure to assess the walking abilities of children undergoing rehabilitation. It is routinely determined during a walking test under standardized conditions, but it remains unclear whether these outcomes reflect the children's performance in daily life. An ankle-worn inertial sensor provides a usable opportunity to measure gait speed in the children's habitual environment. However, sensor-based gait speed estimations need to be accurate to allow for comparison of the children's gait speed between a test situation and daily life. Hence, the first aim of this study was to determine the measurement error of a novel algorithm that estimates gait speed based on data of a single ankle-worn inertial sensor in children undergoing rehabilitation. The second aim of this study was to compare the children’s gait speed between standardized and daily life conditions.MethodsTwenty-four children with walking impairments completed four walking tests at different speeds (standardized condition) and were monitored for one hour during leisure or school time (daily life condition). We determined accuracy by comparing sensor-based gait speed estimations with a reference method in both conditions. Eventually, we compared individual gait speeds between the two conditions.ResultsThe measurement error was 0.01 ± 0.07 m/s under the standardized and 0.04 ± 0.06 m/s under the daily life condition. Besides, the majority of children did not use the same speed during the test situation as in daily life.ConclusionThis study demonstrates an accurate method to measure children's gait speed during standardized walking tests and in the children's habitual environment after rehabilitation. It only requires a single ankle sensor, which potentially increases wearing time and data quality of measurements in daily life. We recommend placing the sensor on the less affected side, unless the child wears one orthosis. In this latter case, the sensor should be placed on the side with the orthosis. Moreover, this study showed that most children did not use the same speed in the two conditions, which encourages the use of wearable inertial sensors to assess the children's walking performance in their habitual environment following rehabilitation.

On-Field Biomechanical Assessment of High and Low Dive in Competitive 16-Year-Old Goalkeepers through Wearable Sensors and Principal Component Analysis.

Diving saves are the main duty of football goalkeepers. Few biomechanical investigations of dive techniques have been conducted, none in a sport-specific environment. The present study investigated the characteristics of goalkeepers’ dive in preferred (PS) and non-preferred (nPS) side through an innovative wearables-plus-principal-component analysis (PCA) approach. Nineteen competitive academy goalkeepers (16.5 ± 3.0 years) performed a series of high and low dives on their PS and nPS. Dives were performed in a regular football goal on the pitch. Full-body kinematics were collected through 17 wearable inertial sensors (MTw Awinda, Xsens). PCA was conducted to reduce data dimensionality (input matrix 310,878 datapoints). PCA scores were extracted for each kinematic variable and compared between PS and nPS if their explained variability was >5%. In high dive, participants exhibited greater hip internal rotation and less trunk lateral tilt (p < 0.047, ES > 0.39) in PS than nPS. In low dives, players exhibited greater ipsilateral hip abduction dominance and lower trunk rotation (p < 0.037, ES > 0.40) in PS than nPS. When diving on their nPS, goalkeepers adopted sub-optimal patterns with less trunk coordination and limited explosiveness. An ecological testing through wearables and PCA might help coaches to inspect relevant diving characteristics and improve training effectiveness.

Technology-based assessment of Short Physical Performance Battery in elderly population

The Short Physical Performance Battery (SPPB) is an objective evaluation of the lower extremity strength [1]. It has been proposed as a reliable method to assess the physical performance of elderly people (65+) through three simple tasks of daily life (gait, standing, chair stand) [2]. In clinical research, the SPPB is administered by research staff who are experienced in the standardized protocol of the examination, with each component measured manually by a stopwatch [3,4]. Therefore, these investigations may be subject to possible interrater and intrarater variabilities [5]. This work aims to validate an automatic method for the objective assessment of SPPB through the combination of wearable inertial and sEMG sensors. The present study included a sub-cohort of Salus in the Apulia Study of 137 old subjects, recruited consecutively from 2019 on a recall-based from an ongoing population-based prospective cohort comprising 2472 individuals aged 65+ and residents in Castellana Grotte, Puglia (Italy). Eligible participants were equipped with an inertial sensor (positioned around the subjets' waist, BTS G- SENSOR2, BTS SpA) and four surface electromyographic probes (BTS FREEEMG, BTS SpA) on Tibialis Anterior (TA) and Gastrocnemius Medialis (GAM) of both legs. The medical doctor instructed subjects to perform three different tasks: balance (stance phase, semi-tandem phase, and tandem phase), five repetition STS and 4-m walk. A stopwatch was used to measure the time required by the subject to complete each task and then assign the SPPB score according to the specific range (SPPB). Kinematics and EMG data of subjects were collected via BTS EMGAnalyzer and exported in MATLAB R2020a to calculate the technology-based SPPB (tSPPB). Descriptive statistics were used to describe the demographic characteristics and outcomes of the sample. The Spearman correlations were used to explore the relationship between the SPPB and tSPPB total score and the three-component scores. We compared the internal consistency of the SPPB and tSPPB using Cronbach’s alpha. Ninety-seven subjects (mean age [±SD] = 75.1 ± 5.53, female = 50.5%) were finally included in the analysis. The total SPPB (9.49 ± 2.55) and tSPPB (8.88 ± 2.51) scores were significantly different, yet highly correlated (r = 0.87, p < 0.01). The alpha coefficient for the SPPB score was high (0.93, 95% CI: 0.90 - 0.96), suggesting that the two methods have relatively high internal consistency. Technology-based assessment of SPPB represents a step forward in adapting the administration of objective measures of physical function to significantly expand the reach of research and clinical practice to assess populations at risk of functional decline.