Human Walking Research Articles

The energetic cost of human walking as a function of uneven terrain amplitude.

Humans expend more energy walking on uneven terrain, but the amount varies across terrains. Few experimental characterizations exist, each describing terrain qualitatively without any relation to others or flat ground. This precludes mechanistic explanation of the energy costs. Here we show that energy cost varies smoothly and approximately quadratically as a function of terrain amplitude. We tested this with healthy adults (N=10) walking on synthetic uneven terrain with random step heights of parametrically controlled maximum amplitude (four conditions 0 - 0.045 m), and at four walking speeds (0.8 - 1.4 m.s-1). Both net metabolic rate and the rate of positive work increased approximately with amplitude squared and speed cubed (R2 = 0.74,0.82 respectively), as predicted by a simple walking model. The model requires work to redirect the body center of mass velocity between successive arcs described by pendulum-like legs, at proportional metabolic cost. Humans performed most of the greater work with terrain amplitude early in the single stance phase, and with speed later in stance during push-off. Work and energy rates changed with approximately linear proportionality, with a ratio or delta efficiency of 49.5% (R2 = 0.68). The efficiency was high enough to suggest substantial work performed passively by elastic tendon and not only by active muscle. Simple kinematic measures such as mid-swing foot clearance also increased with terrain amplitude (R2 = 0.65), possibly costing energy as well. Nevertheless, most of the metabolic cost of walking faster or on more uneven terrain can be explained mechanistically by the work performed.

Characterizing Footstep Detection Thresholds with Distributed Acoustic Sensing

Abstract The resolution of distributed acoustic sensing (DAS) has allowed researchers to detect and analyze a wide variety of seismic and nonseismic sources, such as humans walking. Using preinstalled DAS fibers in telecommunication conduits near human communities makes detecting footstep sources nearly unavoidable. Researchers in the environmental and urban seismology fields see such anthropogenic sources as their signals of interest. However, traditional seismologists may view them as undesirable noise. Targeting or removing such sources requires an understanding of their characteristics. Previous research focused on recording walking signals has demonstrated DAS’s ability to record individual footsteps along a continuous path. Yet, most studies have used footsteps falling immediately above the fiber, with walking paths running parallel to the fiber. Those works provided valuable information on how footstep wavefields propagate. However, previous researchers have not addressed how far away DAS can detect a footstep. Many seismologists use preinstalled DAS with little or no control over array location. Thus, the wider community would benefit from a better understanding of off-fiber footstep detectability. Here, we use wavefields generated by a person walking recorded by a DAS array to locate footsteps through space and time. The footsteps are modeled as a series of discrete impulse forces which are located with backprojection. We explain the recorded amplitudes of a fiber-oblique walking path using a seismic impact model and wave propagation theory. Using this model and the observed amplitudes, we estimate an empirical site-specific maximum footstep detection threshold distance of ≤24 m that is valid for alluvial sites. This threshold can guide researchers in finding ideal DAS array settings, whether they seek to record footstep wavefields or avoid them.

How knee muscles and ground reaction forces shape knee buckling and ankle push-off in neuromuscular simulations of human walking

Ankle push-off is important for efficient, human-like walking, and many prosthetic devices mimic push-off using motors or elastic elements. The knee is extended throughout the stance phase and begins to buckle just before push-off, with timing being crucial. However, the exact mechanisms behind this buckling are still unclear. We use a predictive neuromuscular simulation to investigate whether active muscles are required for knee buckling and to what extent ground reaction forces (GRFs) drive it. In a systematic parameter search, we tested how long the knee muscles vastus (VAS), gastrocnemius (GAS), and hamstrings could be deactivated while maintaining a stable gait with impulsive push-off. VAS deactivation up to 35% of the gait cycle resulted in a dynamic gait with increased ankle peak power. GAS deactivation up to 20% of the gait cycle was detrimental to gait efficiency and showed reduced ankle peak power. At the start of knee buckling, the GRF vector is positioned near the knee joint’s neutral axis, assisting in knee flexion. However, this mechanism is likely not enough to drive knee flexion independently. Our findings contribute to the biomechanical understanding of ankle push-off, with applications in prosthetic and bipedal robotic design, and fundamental research on human gait mechanics.

Kinematics, kinetics, and muscle activations during human locomotion over compliant terrains

Walking on compliant terrains, like carpets, grass, and soil, presents a unique challenge, especially for individuals with mobility impairments. In contrast to rigid-ground walking, compliant surfaces alter movement dynamics and increase the risk of falls. Understanding and modeling gait control across such soft and deformable surfaces is thus crucial for maintaining daily mobility. However, access to the necessary equipment for modeling compliant surface walking is limited. Therefore, in this paper, we present the first publicly available biomechanics dataset of 20 individuals walking on terrains of varying compliance, using a unique robotic device, the Variable Stiffness Treadmill 2 (VST 2), designed to simulate walking on adjustable compliant terrain. VST 2 provides a consistent and reproducible environment for studying the biomechanics of walking on such surfaces within laboratory settings. The goal of this dataset is to provide insights into the muscular, kinematic, and kinetic adaptations that occur when humans walk on compliant terrain in order to design better controllers for prosthetic limbs, improve rehabilitation protocols, and develop adaptive assistive devices that can enhance mobility on compliant surfaces.

Influence of PGV and response spectra on human walking states in simulated earthquake scenarios

Influence of PGV and response spectra on human walking states in simulated earthquake scenarios

Transspinal stimulation downregulates flexion reflex pathways during walking in healthy humans.

The phase-dependent modulation pattern of the tibialis anterior (TA) flexion reflex was characterized during treadmill walking while transspinal stimulation was delivered at 15, 30, and 50 Hz above and below paresthesia in healthy participants. The flexion reflex was elicited following medial arch foot stimulation with a 30 ms (300 Hz) pulse train. During treadmill walking, the flexion reflex was evoked in the right leg every 3 to 5 steps, and stimuli were randomly dispersed across the step cycle that was divided into 16 equal bins. For each participant, condition and bin of the step cycle, the flexion reflex was measured as the area of the linear EMG envelope starting 20 ms after the end of the pulse train up to 200 ms and was normalized to the maximum locomotor TA EMG activity. The unconditioned flexion reflex was modulated in a phase-dependent manner. Transspinal stimulation, regardless frequency, or intensity produced pronounced flexion reflex depression during walking that coincided with an unchanged slope and intercept, computed from the linear relationship between the flexion reflex and background EMG activity. These findings suggest that transspinal stimulation above and below paresthesia intensities at 15, 30, and 50 Hz downregulates the flexion reflex. Based on our recently reported absent effects on the soleus H-reflex under similar conditions and our current findings we propose that transspinal stimulation downregulates flexion and not extension reflex pathways. More research is needed to delineate whether similar neuromodulation effects are present in flexion and extension reflexes after spinal cord injury in humans.

Mechanical Modeling of Plantar Pressure During Human Walking in Different Terrains: Experiments and Analysis

Mechanical Modeling of Plantar Pressure During Human Walking in Different Terrains: Experiments and Analysis

Muscle Control Analysis of Human Walking and Cycling in Speed and Load Variations With Time-Varying Synergy

Muscle Control Analysis of Human Walking and Cycling in Speed and Load Variations With Time-Varying Synergy

Vibration in Floors Induced for Human Walking: Comparison of Two Design Guidelines

The design of floors must fulfill not only the requirements for the Ultimate Limit State but the Serviceability Limit State, which affects, among other factors, the comfort of the users. Due to the modern design of floors and changes in the use of buildings, the effect of human walking and the vibrations induced by this activity can be a determinant factor in the correct design of a floor. National codes are mainly focused on the Ultimate Limit State and give brief recommendations for the Serviceability Limit Sate. In parallel, several design guidelines have been published to assist in the design phase of floors. In this work, we compute the dynamic response of floors excited by human walking using two guidelines, The American Institute of Steel Construction Design Guide (AISC) and the European Guidelines on Human-Induced Vibration of Steel Structures (JRC-ECCS). To compare both design guidelines, we transform the maximum acceleration obtained with the AISC guidelines in velocity and then compare the maximum velocities for different values of natural frequencies, modal masses and damping ratios of the floor under evaluation. The similarities and differences in the response of the floors are studied, as well as the tolerance limits imposed by the guidelines that indicate if the floor design is correct for the comfort of the users.

Modelling individual variation in human walking gait across populations and walking conditions via gait recognition.

Human walking gait is a personal story written by the body, a tool for understanding biological identity in healthcare and security. Gait analysis methods traditionally diverged between these domains but are now merging their complementary strengths to unlock new possibilities. Using large ground reaction force (GRF) datasets for gait recognition is a way to uncover subtle variations that define individual gait patterns. Previously, this was done by developing and evaluating machine learning models on the same individuals or the same dataset, potentially biasing findings towards population samples or walking conditions. This study introduces a new method for analysing gait variation across individuals, groups and datasets to explore how demographics and walking conditions shape individual gait patterns. Machine learning models were implemented using numerous configurations of four large walking GRF datasets from different countries (740 individuals, 7400 samples) and analysed using explainable artificial intelligence tools. Recognition accuracy ranged from 52 to 100%, with factors like footwear, walking speed and body mass playing interactive roles in defining gait. Models developed with individuals walking in personal footwear at multiple speeds effectively recognized novel individuals across populations and conditions (89-99% accuracy). Integrating force platform hardware and gait recognition software could be invaluable for reading the complex stories of human walking.

Characterizing Human Walking Dynamics Across Age: A Control Engineering Approach

Abstract Introduction While various models and studies have investigated the simulation and theoretical description of human gait, previous research has not considered irregular situations such as slipping and tripping. However, understanding these situations could be crucial and provide insight into assessing dynamic balance, a factor that is closely linked to the risk of falling. The first step is to investigate the system under normal walking using a participant group that includes people who are particularly susceptible to an increased risk of falling, such as older adults or people with certain health conditions that have so far been neglected in the literature. Methods In this study a control engineering approach was used to estimate the parameters of the human gait system. A total of 42 subjects aged 18 to 78 years, with different body masses and walking speeds, were analyzed. Measurements were conducted on a treadmill with integrated force plates, while participants wore inertial measurement units (IMUs) at lumbar level and were filmed with cameras. Results The calculated spring-mass-damping model parameters, including stiffness (k) and damping (c), varied widely among participants, with stiffness ranging from 4.45 to 45.53 kN/m and damping from 0.26 to 2.13 kNs/m. Contrary to previous findings, no linear relationships were observed between stiffness or damping and velocity, age, or mass. Outliers were identified and related in a plausible way to visual inspection. Conclusion In conclusion, our study successfully characterized human walking during periodic excitation for a diverse participant group using a simple control engineering approach, yielding comparable results to previous studies. Our results pave the way for future investigations to analyze the response of an individual participant to irregular gait perturbations.

Deep Kronecker LeNet for human motion classification with feature extraction

Human motion classification is gaining more interest among researchers, and it is significant in various applications. Human motion classification and assessment play a significant role in health science and security. Technology-based human motion evaluation deploys motion sensors and infrared cameras for capturing essential portions of human motion and key facial elements. Nevertheless, the prime concern is providing effectual monitoring sensors amidst several stages with less privacy. To overcome this issue, we have developed a human motion categorization system called Deep Kronecker LeNet (DKLeNet), which uses a hybrid network.The system design of impulse radio Ultra-Wide Band (IR-UWB) through-wall radar (TWR) is devised, and the UWB radar acquires the signal. The acquired signal is passed through the gridding phase, and then the feature extraction unit is executed. A new module DKLeNet, which is tuned by Spotted Grey Wolf Optimizer (SGWO), wherein the layers of these networks are modified by applying the Fuzzy concept. In this model, the enhanced technique DKLeNet is unified by Deep Kronecker Network (DKN) and LeNet as well as the optimization modules SGWO is devised by Spotted Hyena Optimizer (SHO) and Grey Wolf Optimizer (GWO). The classified output of human motion is based on human walking, standing still, and empty. The analytic measures of DKLeNet_SGWO are Accuracy, True positive rate (TPR), True Negative rate (TNR), and Mean squared error (MSE) observed as 95.8%, 95.0%, 95.2%, and 38.5%, as well as the computational time observed less value in both training and testing data when compared to other modules with 4.099 min and 3.012 s.

Ventilation strategy for particulate matter control in subway stations

Particulate matter severely affects human health, and its migration is closely related to human walking and ventilation. It is essential to select an appropriate ventilation strategy to control the particulate matter caused by cluster passengers walking. In this study, cluster passengers walk in the narrow passage of a subway station. Large eddy simulation (LES) was used to study the influence of cluster passengers walking on the transmission of particulate matter under ventilation conditions involving upper exhaust and lower exhaust. The risk of passengers exposure was analyzed by comparing the exposure levels of particulate matter at different heights under the effect of human disturbance. The results show that the amount of particulate matter near the legs is greatest during human walking. Clustered passengers disturbance airflow induces particulate matter to rise to the breathing zone, which increases the risk of human exposure. Lower-side exhaust conditions are more conducive to the discharge and control of particulate matter. The disturbance of cluster passengers enhances the continuous transmission of particulate matter in the wake flow. This study is applicable to traffic hubs and other public places where passengers gather, the ventilation of the upper side exhaust has better control effect on particulate matter caused by passengers wake flow.

Exploring the effect of walking patterns on pathway design in landscape architecture using gait analysis

This study investigates the impact of pathway design on human walking patterns using advanced gait analysis techniques to inform landscape architecture. By analyzing key gait parameters such as stride length, cadence, walking speed, step width, and foot placement angles, this research seeks to identify how various pathway features—such as surface material, slope, curvature, and width—influence walking behaviour. Data is collected through motion capture systems and wearable sensors from diverse participants, including individuals of different ages and physical abilities. Statistical methods, including Multivariate Analysis of Variance (MANOVA), are applied to determine significant differences in walking patterns across pathway types, while ML techniques, such as k-means clustering, classify participants based on their walking strategies. The results offer data-driven insights into how different pathway designs affect walking efficiency and comfort. For example, pathways with a slope of 10% reduced WS by 14% compared to flat pathways, while surfaces like gravel increased Foot Placement Angles by 18% compared to concrete, impacting stability. The study provides practical recommendations for creating pathways that support natural human movement, such as ensuring step width and stride length remain consistent across varied surface types by designing smooth transitions between different materials. The study emphasizes the importance of designing inclusive, accessible pathways that accommodate the needs of diverse user groups. For instance, individuals with mobility challenges exhibited a 12% increase in step width on sloped surfaces, suggesting that gentler inclines and smoother textures are essential for accessibility. The findings contribute to LA by offering evidence-based guidelines that optimize pathways’ functionality and user experience in outdoor environments. These guidelines include maintaining a pathway slope below 5% for universal accessibility and using surface materials like concrete or permeable pavers that balance durability and comfort, promoting sustainability and user-centred design.

Center of mass kinematic reconstruction during steady-state walking using optimized template models.

Template models, such as the Bipedal Spring-Loaded Inverted Pendulum and the Virtual Pivot Point, have been widely used as low-dimensional representations of the complex dynamics in legged locomotion. Despite their ability to qualitatively match human walking characteristics like M-shaped ground reaction force (GRF) profiles, they often exhibit discrepancies when compared to experimental data, notably in overestimating vertical center of mass (CoM) displacement and underestimating gait event timings (touchdown/ liftoff). This paper hypothesizes that the constant leg stiffness of these models explains the majority of these discrepancies. The study systematically investigates the impact of stiffness variations on the fidelity of model fittings to human data, where an optimization framework is employed to identify optimal leg stiffness trajectories. The study also quantifies the effects of stiffness variations on salient characteristics of human walking (GRF profiles and gait event timing). The optimization framework was applied to 24 subjects walking at 40% to 145% preferred walking speed (PWS). The findings reveal that despite only modifying ground forces in one direction, variable leg stiffness models exhibited a >80% reduction in CoM error across both the B-SLIP and VPP models, while also improving prediction of human GRF profiles. However, the accuracy of gait event timing did not consistently show improvement across all conditions. The resulting stiffness profiles mimic walking characteristics of ankle push-off during double support and reduced CoM vaulting during single support.

Human walking biomechanics on sand substrates of varying foot sinking depth.

Our current understanding of human gait is mostly based on studies using hard, level surfaces in a laboratory environment. However, humans navigate a wide range of different substrates every day, which incur varied demands on stability and efficiency. Several studies have shown that when walking on natural compliant substrates there is an increase in energy expenditure. However, these studies report variable changes to other aspects of gait such as muscle activity. Discrepancies between studies exist even within substrate types (e.g. sand), which suggests that relatively 'fine-scale' differences in substrate properties exert quantifiable influences on gait mechanics. In this study, we compared human walking mechanics on a range of sand substrates that vary in overall foot sinking depth. We demonstrated that variation in the overall sinking depth in sand was associated with statistically significant changes in joint angles and spatiotemporal variables in human walking but exerted relatively little influence on pendular energy recovery and muscle activations. Significant correlated changes between gait metrics were frequently recovered, suggesting a degree of coupled or mechanistic interaction in their variation within and across substrates. However, only walking speed (and its associated spatiotemporal variables) correlated frequently with absolute foot sinkage depth within individual sand substrates, but not across them. This suggests that a causative relationship between walking speed and foot sinkage depth within individual sand substates is not coupled with systematic changes in joint kinematics and muscle activity in the same way as is observed across sand substrates.

Comparison of Empirical and Reinforcement Learning (RL)-Based Control Based on Proximal Policy Optimization (PPO) for Walking Assistance: Does AI Always Win?

The use of wearable assistive devices is growing in both industrial and medical fields. Combining human expertise and artificial intelligence (AI), e.g., in human-in-the-loop-optimization, is gaining popularity for adapting assistance to individuals. Amidst prevailing assertions that AI could surpass human capabilities in customizing every facet of support for human needs, our study serves as an initial step towards such claims within the context of human walking assistance. We investigated the efficacy of the Biarticular Thigh Exosuit, a device designed to aid human locomotion by mimicking the action of the hamstrings and rectus femoris muscles using Serial Elastic Actuators. Two control strategies were tested: an empirical controller based on human gait knowledge and empirical data and a control optimized using Reinforcement Learning (RL) on a neuromuscular model. The performance results of these controllers were assessed by comparing muscle activation in two assisted and two unassisted walking modes. Results showed that both controllers reduced hamstring muscle activation and improved the preferred walking speed, with the empirical controller also decreasing gastrocnemius muscle activity. However, the RL-based controller increased muscle activity in the vastus and rectus femoris, indicating that RL-based enhancements may not always improve assistance without solid empirical support.

A piezoelectric human motion energy harvester with stress amplification mechanism

The kinetic energy is rich in human walking, which may be exploited to provide sustainable energy for portable and miniaturized devices by properly design of the energy harvester. In order to effectively amplify the actual force on the piezoelectric material upon the external load, a beam-shaped piezoelectric energy harvester (PEH) is designed. The PEH is prepared by embedding poly(vinylidene fluoride) (PVDF) piezoelectric strips into the compression zone of the beam. The underlying mechanism is discussed via mechanical theoretical modeling and finite element simulation. It is demonstrated that the actual integrated force on the piezoelectric film of PVDF is about 31.7 times that of the external force applied on the beam’s mid-span. The study would pave the way for force amplification mechanism in designing PEHs and holds promises for portable device applications.

Multi-branch deep learning neural network prediction model for the development of angular biosensors based on sEMG.

Human gait motion intention recognition is very important for the lower extremity exoskeleton robot to accurately synchronize and respond to the user's natural motion. And motion intention recognition is generally performed through sEMG. Deep learning neural networks perform well in dealing with high-dimensional data and nonlinear relationships such as sEMG, but different deep learning neural networks have their own advantages in dealing with different types of data. Therefore, a multi-branch deep learning neural network, which enables different neural networks to process different feature items, could achieve more accurate and efficient motion intention recognition. The purpose of this study is to 1) Establish a multi-branch deep learning neural network model to achieve accurate gait recognition and effective estimation of joint angles. 2) Quantify the performance of the multi-branch deep learning neural network model in gait recognition and joint angle prediction using sEMG. This study involved the collection of sEMG and plantar pressure data during walking in human subjects. Firstly, the collected signals are filtered and denoised to ensure the quality and reliability of the data. Calculate the time domain features and the frequency domain features to capture the key information of gait. Then, using the sensitivity difference of different structural neural networks to different feature data, a multi-branch deep learning neural network model is developed, in which the extracted features are used as the input of the model. The output of the model includes gait cycle and joint angle, so as to realize the accurate recognition of human gait and the effective estimation of joint angle. The results show that the proposed method has high accuracy in identifying human gait and estimating joint angles. The multi-branch neural network model successfully integrates time-domain and frequency-domain features and provides reliable prediction of gait cycle and joint angle. The highest accuracy of gait recognition is 95.42%, the lowest is 90.11%, and the average is 92.16%. The average error of joint angle estimation is 3.19. This study designed a human walking gait recognition and joint angle prediction model to achieve accurate human lower limb motion intention recognition.The model can be integrated into the sEMG sensor to design a angular biosensors, which can predict the human joint angle in real time.

Theoretical and experimental investigation of flexible air spring stiffness in a tuned liquid column gas damper for vertical vibration control

This study investigates the effectiveness of equivalent linear air spring stiffness in a Vertical Tuned Liquid Column Gas Damper (VTLCGD) for reducing vertical structural vibrations, particularly considering different sealing conditions. The VTLCGD system, designed with a single sealed vertical air column, allows flexible frequency tuning by adjusting the air spring stiffness. It aims to be used in low-frequency structures where conventional VTLCGD designs with two sealed ends are less efficient. The geometric configuration and working principle of the VTLCGD are first described. The equation of motion is derived using liquid dynamic equilibrium and the pressure-volume relationship, with expressions for natural frequency and control force validated through shaking table tests conducted with varying VTLCGD lengths. Experimental investigations are conducted to examine the parameters affecting damper performance using pressure data. The system's vibration reduction performance is then numerically evaluated on a cantilevered floor subjected to human walking. The numerical results, based on the equation of motion for liquid displacement and natural frequency, show good agreement with experimental data, which confirms the effectiveness of using linearized air spring stiffness for frequency tuning. The effects of total liquid length and height difference on control force are further investigated, and a polytropic index of 1.2 is determined for the VTLCGD frequency formula. The VTLCGD system achieves a maximum vibration reduction of 52.63 % in acceleration response on the cantilevered floor compared to the uncontrolled case. Moreover, the variation in stiffness due to changes in the sealing condition is presented to validate the damper's adaptability over a wide frequency range.