Human Activity Monitoring Research Articles

Wearable Sensors for Physiological Condition and Activity Monitoring

Rapid technological advancements have transformed the healthcare sector from traditional diagnosis and treatment to personalized health management. Biofluids such as teardrops, sweat, interstitial fluids, and exhaled breath condensate offer a rich source of metabolites that can be linked to the physiological status of an individual. More importantly, these biofluids contain biomarkers similar to those in the blood. Therefore, developing sensors for the noninvasive determination of biofluid‐based metabolites can overcome traditionally invasive and laborious blood‐test‐based diagnostics. In this context, wearable devices offer real‐time and continuous physiological conditions and activity monitoring. The first‐generation wearables included wristwatches capable of tracking heart rate variations, breathing rate, body temperature, stress responses, and sleeping patterns. However, wearable sensors that can accurately measure the metabolites are needed to achieve real‐time analysis of biomarkers. In this review, recent progresses in wearable sensors utilized to monitor metabolites in teardrops, breath condensate, sweat, and interstitial fluids are thoroughly analyzed. More importantly, how metabolites can be selectively detected, quantified, and monitored in real‐time is discussed. Furthermore, the review includes a discussion on the utility of, multifunctional sensors that combine metabolite sensing, human activity monitoring, and on‐demand drug delivery system for theranostic applications.

A mathematical model to design pillar-shaped bioinspired mechanical sensors

Soft robots and sensor/actuator systems are often based on bioinspired designs to leverage nature patterns. Specifically, pillar-shaped sensors are useful for human activity monitoring, locomotion of soft robots or treatment of cardiovascular diseases. If electric or magnetic particles are added in the manufacturing process, these structures can be tuned through remote fields to attain a specific mechanical behaviour. This promising technique has direct applications in high-impact fields such as bioengineering, soft robots or sensor designing. Filament-shaped smart sensors can send electrical signals when subjected to an external mechanical stimulus or provide a mechanical response when a remote and controllable field is applied broadening their possibilities of action. As the efficient design of these structures is highly challenging, developing a technical tool with a low computational cost to help throughout layout processes (i.e. inverse engineering) is pivotal. Theoretical modelling of the kinematics and dynamics of a wire-shaped structure under an external action is the first step to provide a methodology to help designing mechanical sensors in an efficient, understandable and low time-consuming way. The event of mechanical deformation after receiving the external stimulus and before sending the corresponding output signal is key in the conceptualisation process of smart sensors. This work intends to give insight into the dynamics of a deformable pillar-shaped sensor component under an external action without addressing or coupling its causes and, hence, provide the general mechanical framework to serve as basis for multiphysics formulations for pillar-shaped sensor design.

Tunable porous fiber-shaped strain sensor with synergistic conductive network for human motion recognition and tactile sensing

With the rapid development of portable electronic products, smart wearable devices show great application potential in personalized motion monitoring, health monitoring and even in human-machine interaction. In particular, high-performance flexible fiber-shaped strain sensors are receiving more and more attention because they can be well integrated with clothes or human skin for real-time human activities monitoring. However, achieving improved sensitivity, wide detection range, diverse applications and easy operation of them remains a challenge. Here, we adopt a simple wet-spinning technique to prepare a porous fiber-shaped carbon nanotubes (CNTs)/silver nanoparticles (AgNPs)/thermoplastic polyurethane (TPU) fiber (CATF) for high-performance strain sensor, of which the synergistic conductive network of CNTs and AgNPs enables a wide detection range (∼248 %), high sensitivity (GF > 4295), fast response/recovery time (120 ms/140 ms) and excellent stability. Interestingly, strain sensing range and sensitivity of the sensor can be tuned by changing the AgNPs loading to make it suitable for different application scenarios. All the merits of the sensor make it capable for diverse human movements monitoring, pronunciation recognition, and gesture recognition. More importantly, our prepared CATF strain sensor can be easily weaved into wearable smart textiles, showing great potential for tactile sensing to realize multi-touch and pressure tracking. Finally, the CATF also displays efficient photothermal conversion capability even under different tensile strains up to 150 %, showing great potential in human body thermal management to improve the wearing comfort of wearable electronic fabric.

Multi-head CNN-based activity recognition and its application on chest-mounted sensor-belt

In recent years, a great deal of research has been done on the identification, monitoring, and classification of human activities. Human activity recognition (HAR) is a term commonly used to describe the automatic identification of physical activities. For activity recognition, there are primarily vision-based and sensor-based methods available. The computer vision-based method is generally effective in lab settings, but because of clutter, fluctuating light levels, and contrast, it may not perform well in real-world scenarios. Continuous monitoring and analysis of physiological signals obtained from heterogeneous sensors attached to an individual’s body is required to realise sensor-based HAR systems. Most of the previous research in human activity recognition (HAR) is biased along with feature engineering and pre-processing which requires a good amount of domain knowledge. Application-specific modelling and time-taking methods are involved in these approaches. In this work, the multi-head convolutional neural network-based human activity recognition framework is proposed where automatic feature extraction and classification are involved in the form of an end-to-end classification approach. Experiments of this approach are performed by taking raw wearable sensor data with few pre-processing steps and without the involvement of a handcrafted feature extraction technique. 99.23% and 93.55% accuracy are obtained on the WISDM and UCI-HAR datasets which denoted the much improvement in the assessment of HAR over other similar approaches. The model is also tested on locally collected data from a chest mounted belt with fabric sensors and an accuracy of 87.14% has been achieved on that data.

CQDs-Cross-Linked Conductive Collagen/PAA-Based Nanocomposite Organohydrogel Coupling Flexibility with Multifunctionality for Dual-Modal Sensing of Human Motions.

Conductive hydrogels are ideal materials for intelligent medical devices, human-machine interfaces, and flexible bioelectrodes due to their adjustable mechanical properties and electrical responsiveness, whereas it is still a great challenge to achieve the integration of excellent flexibility and biocompatibility into one hydrogel sensor while also incorporating self-healing, self-adhesion, environmental tolerance, and antimicrobial properties. Here, a nanocomposite conductive organohydrogel was constructed by using collagen (Col), alginate-derived carbon quantum dots (OSA-CQDs), poly(acrylic acid) (PAA), ethylene glycol reduced AgNPs, and Fe3+ ions. Depending on OSA-CQDs with multiple chemical binding sites and high specific surface area as cross-linkers, while coupling highly biologically active Col chains and PAA chains are serving as an energy dissipation module, the resulting organohydrogel exhibited excellent flexibility (795% of strain, 193 kPa of strength), high cell compatibility (>95% survival rate), self-healing efficiency (HE = 79.5%), antifreezing (-20 °C), moisturizing (>120 h), repeatable adhesion (strength >20 kPa, times >10), inhibitory activity against Escherichia coli and Staphylococcus aureus (9 and 21.5 cm2), conductivity, and strain sensitivity (σ = 1.34 S/m, gauge factor (GF) = 11.63). Based on the all-in-one integration of multifunction, the organohydrogel can collaboratively adapt to the multimode of strain sensing and electrophysiological sensing to realize wireless real-time monitoring of human activities and physiological health. Therefore, this work provides a new and common platform for the design and sensing of next-generation hydrogel-based smart wearable sensors.

All porous Ecoflex and SEBS-based stretchable high-performance triboelectric nanogenerator for self-powered human activity monitoring

Achieving a high output performance by expanding the surface area of triboelectric materials is key to developing triboelectric nanogenerators (TENGs) for future versatile applications. Herein, highly porous Ecoflex and styrene-ethylene-butylene-styrene (SEBS) using chloroform are introduced to improve TENG performance. The triboelectric output increases significantly with increasing porosity, which can be attributed to the increase in the contact area, dielectric constant, and electrostatic induction in the porous structure, thereby resulting in additional surface charges on the porous surface and increased surface charge density of the friction materials. The TENG can stretch by up to 550 % with the designed fabric electrode configuration. Facilitated by the increased contact area of 3D printed microstructure and porous Ecoflex substrate, the fabricated TENG demonstrates a high voltage of 2280 V, corresponding to a power density of 18.8 W m−2, which is enough to power 460 light-emitting diodes and stopwatches. Demonstration of monitoring the walking steps of patients with Parkinson’s and blind people was also performed, which revealed outstanding sensing capabilities. In addition, the TENG-based self-powered sensor demonstrated an identification accuracy of 99.3 % and a motion detection accuracy of 98.2 % based on a deep learning model. Furthermore, by exploiting the outstanding ultrahigh sensitivity, the TENG was demonstrated for low-frequency football player motion monitoring. Therefore, the results of this study provide new insights into investigating porous materials and an efficient method to construct high-performance TENGs that can promote their practical application in the fields of smart healthcare and sports.

Human Eye Activity Monitoring Using Continuous Wave Doppler Radar: A Feasibility Study.

Human eye activity has been widely studied in many fields such as psychology, neuroscience, medicine, and human-computer interaction engineering. In previous studies, monitoring of human eye activity mainly depends on electrooculogram (EOG) that requires a contact sensor. This article proposes a novel eye movement monitoring method called continuous wave doppler oculogram (cDOG). Unlike the conventional EOG-based eye movement monitoring methods, cDOG based on continuous wave doppler radar sensor (cDRS) can remotely measure human eye activity without placing electrodes on the head. To verify the feasibility of using cDOG for eye movement monitoring, we first theoretically analyzed the association between the radar signal and the corresponding eye movements measured with EOG. Afterward, we conducted an experiment to compare EOG and cDOG measurements under the conditions of eyes closure and opening. In addition, different eye movement states were considered, including right-left saccade, up-down saccade, eye-blink, and fixation. Several representative time domain and frequency domain features obtained from cDOG and from EOG were compared in these states, allowing us to demonstrate the feasibility of using cDOG for monitoring eye movements. The experimental results show that there is a correlation between cDOG and EOG in the time and frequency domain features, the average time error of single eye movement is less than 280.5 ms, and the accuracy of cDOG in eye movement detection is higher than 92.35%, when the distance between the cDRS and the face is 10 cm and eyes is facing the radar directly.

MXene-Based Pressure Sensor with a Self-Healing Property for Joule Heating and Friction Sliding.

As a kind of flexible electronic device, flexible pressure sensor has attracted wide attention in medical monitoring and human-machine interaction. With the continuous deepening of research, high-sensitivity sensor is developing from single function to multi-function. However, Current multifunctional sensors lack the ability to integrate joule heating, detect sliding friction, and self-healing. Herein, a MXene/polyurethane (PU) flexible pressure sensor with a self-healing property for joule heating and friction sliding is fabricated. The MXene/PU sensitive layer with special spinosum structure is prepared by a simple spraying method. After face-to-face assembly of the sensitive layers, the MXene/PU flexible pressure sensor is obtained and showed excellent sensitivity (150.65kPa-1), fast response/recovery speed (75.5/63.9ms), and good stability (10000 cycles). Based on the self-healing property of PU, the sensor also has the ability to heal after mechanical damage. In addition, the sensor realizes the joule heating function under low voltage, and has the real-time monitoring ability of sliding objects. Combined with low cost and simple manufacturing method, the multi-functional MXene/PU flexible sensor shows a wide range of application potential in human activity monitoring, thermal management, and slip recognition.

A spatio-temporal graph neural network for fall prediction with inertial sensors

Falls are the leading cause of unintentional human injury, having become a public health event of strong social concern. The fall prediction technology based on wearable inertial sensors is a relatively reliable solution in human activity monitoring, a user scenario with mobility and high information privacy sensitivity, and has the advantages of low cost, small size, and high precision. However, a key challenge in sensor-based fall prediction is the fact that a fall event can often occur in various configurations of fall poses together with their own spatio-temporal dependencies. This leads us to define a spatio-temporal model to explicitly characterize these internal configurations of poses. In particular, we introduce a graph neural network with spatio-temporal topological structure to encode such latent relations among poses by capturing representative patterns in fall events. Moreover, a human body orientation estimator is devised to represent human low limbs information and as a result, separate pose dependencies are globally consistent. Empirical evaluations on two benchmark datasets and one in-house dataset suggest our approach significantly outperforms the state-of-the-art methods.

Three-dimensional surface strain sensor based on PDMS/LIG composite film with adjustable electromechanical performance

Flexible strain sensors play a crucial role in health monitoring, smart wearable devices, and human–machine interaction. Three-dimensional surface evaluation methods for strain sensors offer advantages by being closer to actual strain, featuring a larger working range, and being more suitable for multidirectional strain. In this study, a three-dimensional (3D) surface strain sensor based on polydimethylsiloxane/laser-induced graphene (PDMS/LIG) composite films has been developed. The electromechanical properties of this sensor, encompassing 3D strain range and sensitivity, can be adjusted by manipulating laser parameters and LIG patterns. The key to attaining these specific characteristics lies in the intentional design of crack types and orientations on the sensor's surface. Remarkably, the line-vertical (LV) sensor exhibits outstanding sensitivity with a GF of 211.3. The line-parallel (LP) sensor achieves a GF of 115.1. Additionally, it demonstrates a stretching range of 25% and maintains stable performance over an extensive number of strain/release test cycles (more than 3000 cycles). With these advantages, the 3D strain sensor can not only be applied in human activity monitoring but also monitoring pressure within microchannels in microfluidic chips, suggesting promising applications in the health and medical fields.

Metal-free flexible triboelectric nanogenerator based on bifunctional carbon fiber for mechanical energy harvesting and human activity monitoring

Conductive electrodes in triboelectric nanogenerators (TENGs) typically use metal-based materials. However, such devices have limited flexibility and challenges in humid environments. To overcome this issue, we present a metal-free flexible TENG made from bifunctional carbon fiber paper that serves as both a conductive electrode and an efficient tribo-positive friction layer. The electrical performance of carbon fiber paper-based TENG (CFP-TENG) at 3 Hz demonstrated a peak-to-peak voltage of 109.1 V, a current of 10.8 μA, and a power density of ∼ 0.093 W/m2 at 7 MΩ external resistance. A 20 µF capacitor is also charged, and 38 light-emitting diodes (LEDs) arranged in the letter 'DGU' can be lighted. The CFP-TENG device's high flexibility allows it to sustain 12,000 repeated mechanical contact-separation cycles and fold, twist, and roll the device without degrading performance. In addition, the hydrophobic characteristic of carbon fiber ensures the improved moisture resistance of the CFP-TENG device compared to conventional copper electrodes. Furthermore, the CFP-TENG is successfully integrated into various human joints as a self-powered motion sensor for tracking real-time human activity monitoring, including finger motion detection, arm motion, walking, and running activities. A metal-free TENG based on bifunctional carbon fibers paves the way for energy- harvesting solutions and self-powered biomedical sensors.

Environment‐Tolerant Conductive Eutectogels for Multifunctional Sensing

AbstractConductive hydrogels have shown significant potential in the realm of flexible electronics; however, using water as a solvent introduces problems such as low‐temperature freezing and water loss. One possible solution to address these limitations is replacing water with deep eutectic solvents (DES). In this study, the PHEAA–gelatin–MXene (PGM) eutectogels are constructed with poly(N‐hydroxyethyl acrylamide) (PHEAA) and gelatin as the main components, and MXene nanosheets are added as nanofillers. The PGM eutectogels exhibit exceptional tensile and compressive mechanical properties, including remarkable stretchability (940%), high strength (0.5 MPa), high toughness (1.39 MJ m−3), and impressive compressive strength (0.3 MPa at 80% strain). Additionally, the PGM eutectogels demonstrate excellent adhesion, anti‐freezing, and long‐term anti‐drying abilities. Moreover, the PGM eutectogels multifunctional sensors allow high sensitivity, which enables accurate real‐time and stable monitoring of human activities over a wide temperature range. Consequently, the PGM eutectogels hold great potential as candidates in fields such as flexible wearable devices, personal healthcare, and human–machine interfaces.

Stretchable and Highly Sensitive Flexible Strain Sensor Based on a Three‐Layer Core–Shell Structure of Polydopamine/Polypyrrole@Natural Rubber for Human Activity Monitoring

In recent years, flexible strain sensors with a wide detection range and high sensitivity have garnered significant attention for human health monitoring. However, fabricating strain sensors with a broad strain range while retaining high sensitivity remains a major challenge. Herein, a green and facile method is successfully developed to fabricate the polydopamine/polypyrrole@natural rubber (PDA/PPy@NR)‐based strain sensor with a three‐layer core–shell structure. PDA and PPy are coated uniformly onto the surface of the NR nanosphere by in situ polymerization. The electromechanical and strain sensing properties of the fabricated composite are investigated. The homogeneous coating of PPy and PDA layers has facilitated the formation of continuous conductive pathways within the NR matrix, even at low loading levels. Consequently, the strain sensor exhibits remarkable sensitivity over a wide strain range (≈800%) and long‐term reliability (2500 cycles at 50%). The PDA/PPy@NR strain sensor shows an exceptionally high gauge factor value (142.6) and a short response time (125 ms). Furthermore, this PDA/PPy@NR strain sensor can effectively detect and capture subtle and large human movements such as limb joints and facial movements as well as even distinguish the pronunciation of different words. The PDA/PPy@NR strain sensor holds great promise for integration into flexible wearable electronics.

Semantic representation and comparative analysis of physical activity sensor observations using MOX2-5 sensor in real and synthetic datasets: a proof-of-concept-study

The widespread use of devices like mobile phones and wearables allows for automatic monitoring of human daily activities, generating vast datasets that offer insights into long-term human behavior. A structured and controlled data collection process is essential to unlock the full potential of this information. While wearable sensors for physical activity monitoring have gained significant traction in healthcare, sports science, and fitness applications, securing diverse and comprehensive datasets for research and algorithm development poses a notable challenge. In this proof-of-concept study, we underscore the significance of semantic representation in enhancing data interoperability and facilitating advanced analytics for physical activity sensor observations. Our approach focuses on enhancing the usability of physical activity datasets by employing a medical-grade (CE certified) sensor to generate synthetic datasets. Additionally, we provide insights into ethical considerations related to synthetic datasets. The study conducts a comparative analysis between real and synthetic activity datasets, assessing their effectiveness in mitigating model bias and promoting fairness in predictive analysis. We have created an ontology for semantically representing observations from physical activity sensors and conducted predictive analysis on data collected using MOX2-5 activity sensors. Until now, there has been a lack of publicly available datasets for physical activity collected with MOX2-5 activity monitoring medical grade (CE certified) device. The MOX2-5 captures and transmits high-resolution data, including activity intensity, weight-bearing, sedentary, standing, low, moderate, and vigorous physical activity, as well as steps per minute. Our dataset consists of physical activity data collected from 16 adults (Male: 12; Female: 4) over a period of 30–45 days (approximately 1.5 months), yielding a relatively small volume of 539 records. To address this limitation, we employ various synthetic data generation methods, such as Gaussian Capula (GC), Conditional Tabular General Adversarial Network (CTGAN), and Tabular General Adversarial Network (TABGAN), to augment the dataset with synthetic data. For both the authentic and synthetic datasets, we have developed a Multilayer Perceptron (MLP) classification model for accurately classifying daily physical activity levels. The findings underscore the effectiveness of semantic ontology in semantic search, knowledge representation, data integration, reasoning, and capturing meaningful relationships between data. The analysis supports the hypothesis that the efficiency of predictive models improves as the volume of additional synthetic training data increases. Ontology and Generative AI hold the potential to expedite advancements in behavioral monitoring research. The data presented, encompassing both real MOX2-5 and its synthetic counterpart, serves as a valuable resource for developing robust methods in activity type classification. Furthermore, it opens avenues for exploration into research directions related to synthetic data, including model efficiency, detection of generated data, and considerations regarding data privacy.

Multifunctional, Degradable Wearable Sensors Prepared with an Initiator and Crosslinker-Free Method.

The present zwitterionic hydrogel-based wearable sensor exhibits various limitations, such as limited degradation capacity, unavoidable toxicity resulting from initiators, and poor mechanical properties that cannot satisfy practical demands. Herein, we present an initiator and crosslinker-free approach to prepare polyethylene glycol (PEG)@poly[2-(methacryloyloxy)ethyl] dimethyl-(3-sulfopropyl) (PSBMA) interpenetrating polymer network (IPN) hydrogels that are self-polymerized via sunlight-induced and non-covalent crosslinking through electrostatic interaction and hydrogen bonding among polymer chains. The PEG@PSBMA IPN hydrogel possesses tissue-like softness, superior stretchability (∼2344.6% elongation), enhanced fracture strength (∼39.5 kPa), excellent biocompatibility, antibacterial property, reliable adhesion, and ionic conductivity. Furthermore, the sensor based on the IPN hydrogel demonstrates good sensitivity and cyclic stability, enabling effective real-time monitoring of human body activities. Moreover, it is worth noting that the excellent degradability in the saline solution within 8 h makes the prepared hydrogel-based wearable sensor free from the electronic device contamination. We believe that the proposed strategy for preparing physical zwitterionic hydrogels will pave the way for fabricating eco-friendly wearable devices.

Large strain, tissue-like and self-healing conductive double-network hydrogel for underwater information transmission

Traditional information-transmission systems are poorly suited for rapid underwater communication because of their large size, high cost, and low flexibility, thereby inspiring the development of conductive hydrogel-based alternatives. However, the low extensibility and non-tissue-like flexibility of common hydrogels limit their applications in wearable devices for underwater use. As such, a self-healing double-network hydrogel with a large maximal strain (>1200 %), high conductivity (3.61 S m−1) and fatigue resistance, and tissue-like Young’s modulus (<20.4 kPa) was prepared by freeze–thaw and solvent-replacement methods using polyvinyl alcohol and sodium alginate as the raw materials. In addition, the suitability of the resulting self-healing hydrogel for land and underwater applications, including sensing, communication, and energy collection, was demonstrated. The developed material enabled the (i) effective monitoring of multiple human physiological activities on land and (ii) highly consistent real-time underwater information transmission based on the use of the international Morse code at reasonable water temperatures (0–25 °C) and frequencies (1–5 Hz). Thus, this study presents a strategy for the development of soft materials as components of intelligent underwater electronic devices.

Simultaneously enhancing sensitivity and operation range of flexible pressure sensor by constructing a magnetic-guided microstructure in laser-induced graphene composite

Flexible pressure sensors play an essential role in the development of flexible electronic systems, which have been widely used in recent times for human activity monitoring and personal healthcare. However, possessing both high sensitivity and a wide operation range for pressure sensor remains challenging, as these requirements are commonly incompatible with each other. Herein, a magnetic induced graphene composite-based pressure sensor (MIPS) is demonstrated by integrating laser induced graphene (LIG) with a patterned valley-peak microstructure and a graphene film doped with magnetic microspheres. Owing to the synergistic effect of the 3D valley-peak microstructures and intercalation-like microstructures within the LIG composite film, the MIPS demonstrates excellent sensing performance in both sensitivity and operation range. The conductive network formed by graphene flakes improves electromechanical response capability, while the incorporation of magnetic microspheres (MMs) promotes the proliferation of conductive pathways in a small applied pressure range, and the valley-peak structure further enhances compressive strength over a large operation range. Accordingly, the MIPS exhibit excellent sensing capability over a range of 0–830 kPa, with an exceptionally high sensitivity of 4426.12 kPa−1 and excellent mechanical stability and durability (2000 cycles at a pressure of 800 kPa). Additionally, we have developed a novel pressure mapping system utilizing the MIPS array, enabling real-time monitoring of pressure spatial distribution and intensity. Notably, the MIPS array exhibits remarkable potential in diverse applications, including the monitoring of human physiological signals and movement. This pressure sensor with a hierarchically composite microstructure presents a promising candidate for the rapid advancement of next-generation wearable bioelectronics.

A dynamically stable self-healable wire based on mechanical-electrical coupling.

The rise in wearable electronics has witnessed the advancement of self-healable wires, which are capable of recovering mechanical and electrical properties upon structural damage. However, their highly fluctuating electrical resistances in the range of hundreds to thousands of ohms under dynamic conditions such as bending, pressing, stretching and tremoring may seriously degrade the precision and continuity of the resulting electronic devices, thus severely hindering their wearable applications. Here, we report a new family of self-healable wires with high strengths and stable electrical conductivities under dynamic conditions, inspired by mechanical-electrical coupling of the myelinated axon in nature. Our self-healable wire based on mechanical-electrical coupling between the structural and conductive components has significantly improved the electrical stability under dynamic scenarios, enabling precise monitoring of human health status and daily activities, even in the case of limb tremors from simulated Parkinson's disease. Our mechanical-electrical coupling strategy opens a new avenue for the development of dynamically stable electrodes and devices toward real-world wearable applications.

Intelligent Millimeter-Wave System for Human Activity Monitoring for Telemedicine.

Telemedicine has the potential to improve access and delivery of healthcare to diverse and aging populations. Recent advances in technology allow for remote monitoring of physiological measures such as heart rate, oxygen saturation, blood glucose, and blood pressure. However, the ability to accurately detect falls and monitor physical activity remotely without invading privacy or remembering to wear a costly device remains an ongoing concern. Our proposed system utilizes a millimeter-wave (mmwave) radar sensor (IWR6843ISK-ODS) connected to an NVIDIA Jetson Nano board for continuous monitoring of human activity. We developed a PointNet neural network for real-time human activity monitoring that can provide activity data reports, tracking maps, and fall alerts. Using radar helps to safeguard patients' privacy by abstaining from recording camera images. We evaluated our system for real-time operation and achieved an inference accuracy of 99.5% when recognizing five types of activities: standing, walking, sitting, lying, and falling. Our system would facilitate the ability to detect falls and monitor physical activity in home and institutional settings to improve telemedicine by providing objective data for more timely and targeted interventions. This work demonstrates the potential of artificial intelligence algorithms and mmwave sensors for HAR.

Leveraging Smartphone Sensor Data and Machine Learning Model for Human Activity Recognition and Fall Classification

This study investigates the application of Machine Learning (ML) algorithms to classify human physical activities and detect different types of falls using smartphone sensor data. Smartphones equipped with various sensors offer new opportunities for healthcare and human activity monitoring, which are crucial for health and well-being, aligning with the United Nations’ sustainable development goals (SDG). The study used the ”Sensor Data Collector” Android application to collect smartphone sensor data. The data collection experiments involve ten healthy participants engaged in diverse physical activities and different fall events, utilizing accelerometers and gyroscope sensors. Different types of fall detection include forward, backward, left lateral, and right lateral fall, which are important from a seriousness of health point of view. Participants placed their smartphones in the front pocket of their trousers to ensure a fixed position during the data collection process. The collected datasets are preprocessed using a moving average filter to denoise and smoothen the data. Different statistical features and extracted and ML algorithms, namely Support Vector Machine (SVM), Random Forest (RF), Naive Bayes (NB), and K-Nearest Neighbors (KNN), are employed to create robust classifiers. SVM and RF performed exceptionally well in classifying normal physical activities, achieving a 99.09% accuracy. SVM excelled in daily walking activity classification with a 99.59% accuracy, while the K-NN model performed best for crucial different types of fall classification, achieving an accuracy of 85.71%. This research emphasizes the potential of ML models for accurate physical activity and various types of fall classification, expanding the horizons of smartphone-based healthcare applications and aligning with sustainable development goals.