Triaxial Gyroscope Research Articles

Adaptive backstepping control design for MEMS gyroscope based on function approximation techniques with input saturation and output constraints

An efficient adaptive backstepping control method is presented in the current study for controlling the micro-electro-mechanical system (MEMS) triaxial gyroscope. In this approach, practical considerations, including output constraints and input saturation, are taken into account. A Barrier Lyapunov Function (BLF) is used so that non-transgression of the constraints is ensured. In the suggested control method, for approximating the lumped uncertainties and addressing the problem of explosion of complexity, Hermite polynomials are used as a simple but effective function approximation approach as a universal approximator. Therefore, the resulting control approach is much simpler with minimal learning parameters. The Lyapunov stability theory is used so that the semiglobally uniformly ultimately boundedness of the closed-loop signals with proper tracking error convergence is guaranteed. According to the results of the comparative simulation, the superiority of this approach is confirmed.

A Measurement of ‘Walking-the-Wall’ Dynamics: An Observational Study Using Accelerometry and Sensors to Quantify Risk Associated with Vertical Wall Impact Attenuation in Trampoline Parks

This study illustrates the application of a tri-axial accelerometer and gyroscope sensor device on a trampolinist performing the walking-the-wall manoeuvre on a high-performance trampoline to determine the performer dynamic conditions. This research found that rigid vertical walls would allow the trampolinist to obtain greater control and retain spatial awareness at greater levels than what is achievable on non-rigid vertical walls. With a non-rigid padded wall, the reaction force from the wall can be considered a variable force that is not constrained, and would not always provide the feedback that the trampolinist needs to maintain the balance with each climb up the wall and fall from height. This research postulates that unattenuated vertical walls are safer than attenuated vertical walls for walking-the-wall manoeuvres within trampoline park facilities. This is because non-rigid walls would provide higher g-force reaction feedback from the wall, which would reduce the trampolinist’s control and stability. This was verified by measuring g-force on a horizontal rigid surface versus a non-rigid surface, where the g-force feedback was 27% higher for the non-rigid surface. Control and stability are both critical while performing the complex walking-the-wall manoeuvre. The trampolinist experienced a very high peak g-force, with a maximum g-force of approximately 11.5 g at the bottom of the jump cycle. It was concluded that applying impact attenuation padding to vertical walls used for walking-the-wall and similar activities would increase the likelihood of injury; therefore, padding of these vertical surfaces is not recommended.

Monolithic Fully Decoupled Tri-Axis Gyroscope With Lateral-Comb and Low Cross-Coupling

Multi-axis integration is important for micro-electromechanical system(MEMS) gyroscopes in the future. In this paper, a monolithic integrated tri-axis gyroscope and interface circuit are designed and implemented. The presented tri-axis gyroscope with single drive and multi sense axes is fully decoupled in structure among the drive and sense modes. Quadrature coupling can also be further reduced by the in-plane and out-of-plane correction electrodes. The pitch(roll) structure is designed as a unilateral unequal height comb, which reduces the damping and improves the Q value of pitch and roll modes and simplifies the fabrication process. In addition, the interface circuit adopts a closed-loop drive and open-loop sense circuit, an additional decoupling module that effectively reduces the cross axis sensitivity. Finally, a gyroscope prototype was tested to evaluate its performance. The designed tri-axis gyroscope has low bias instability(1.2821°/h) and good noise performance(0.1635°/ <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\surd \text{h}$ </tex-math></inline-formula> ). The coupling coefficient is even below 1% after adding decoupling modules. Meanwhile, the bandwidth can meet the needs of the gyroscope working in a low-frequency environment.

Deep Learning-Based Optimal Smart Shoes Sensor Selection for Energy Expenditure and Heart Rate Estimation.

Wearable technologies are known to improve our quality of life. Among the various wearable devices, shoes are non-intrusive, lightweight, and can be used for outdoor activities. In this study, we estimated the energy consumption and heart rate in an environment (i.e., running on a treadmill) using smart shoes equipped with triaxial acceleration, triaxial gyroscope, and four-point pressure sensors. The proposed model uses the latest deep learning architecture which does not require any separate preprocessing. Moreover, it is possible to select the optimal sensor using a channel-wise attention mechanism to weigh the sensors depending on their contributions to the estimation of energy expenditure (EE) and heart rate (HR). The performance of the proposed model was evaluated using the root mean squared error (RMSE), mean absolute error (MAE), and coefficient of determination (). Moreover, the RMSE was 1.05 ± 0.15, MAE 0.83 ± 0.12 and 0.922 ± 0.005 in EE estimation. On the other hand, and RMSE was 7.87 ± 1.12, MAE 6.21 ± 0.86, and 0.897 ± 0.017 in HR estimation. In both estimations, the most effective sensor was the z axis of the accelerometer and gyroscope sensors. Through these results, it is demonstrated that the proposed model could contribute to the improvement of the performance of both EE and HR estimations by effectively selecting the optimal sensors during the active movements of participants.

Monitoring Flexions and Torsions of the Trunk via Gyroscope-Calibrated Capacitive Elastomeric Wearable Sensors.

Reliable, easy-to-use, and cost-effective wearable sensors are desirable for continuous measurements of flexions and torsions of the trunk, in order to assess risks and prevent injuries related to body movements in various contexts. Piezo-capacitive stretch sensors, made of dielectric elastomer membranes coated with compliant electrodes, have recently been described as a wearable, lightweight and low-cost technology to monitor body kinematics. An increase of their capacitance upon stretching can be used to sense angular movements. Here, we report on a wearable wireless system that, using two sensing stripes arranged on shoulder straps, can detect flexions and torsions of the trunk, following a simple and fast calibration with a conventional tri-axial gyroscope on board. The piezo-capacitive sensors avoid the errors that would be introduced by continuous sensing with a gyroscope, due to its typical drift. Relative to stereophotogrammetry (non-wearable standard system for motion capture), pure flexions and pure torsions could be detected by the piezo-capacitive sensors with a root mean square error of ~8° and ~12°, respectively, whilst for flexion and torsion components in compound movements, the error was ~13° and ~15°, respectively.

Classifying Goliath Grouper (Epinephelus itajara) Behaviors from a Novel, Multi-Sensor Tag.

Inertial measurement unit sensors (IMU; i.e., accelerometer, gyroscope and magnetometer combinations) are frequently fitted to animals to better understand their activity patterns and energy expenditure. Capable of recording hundreds of data points a second, these sensors can quickly produce large datasets that require methods to automate behavioral classification. Here, we describe behaviors derived from a custom-built multi-sensor bio-logging tag attached to Atlantic Goliath grouper (Epinephelus itajara) within a simulated ecosystem. We then compared the performance of two commonly applied machine learning approaches (random forest and support vector machine) to a deep learning approach (convolutional neural network, or CNN) for classifying IMU data from this tag. CNNs are frequently used to recognize activities from IMU data obtained from humans but are less commonly considered for other animals. Thirteen behavioral classes were identified during ethogram development, nine of which were classified. For the conventional machine learning approaches, 187 summary statistics were extracted from the data, including time and frequency domain features. The CNN was fed absolute values obtained from fast Fourier transformations of the raw tri-axial accelerometer, gyroscope and magnetometer channels, with a frequency resolution of 512 data points. Five metrics were used to assess classifier performance; the deep learning approach performed better across all metrics (Sensitivity = 0.962; Specificity = 0.996; F1-score = 0.962; Matthew’s Correlation Coefficient = 0.959; Cohen’s Kappa = 0.833) than both conventional machine learning approaches. Generally, the random forest performed better than the support vector machine. In some instances, a conventional learning approach yielded a higher performance metric for particular classes (e.g., the random forest had a F1-score of 0.971 for backward swimming compared to 0.955 for the CNN). Deep learning approaches could potentially improve behavioral classification from IMU data, beyond that obtained from conventional machine learning methods.

An Efficient Calibration Method for Triaxial Gyroscope

This paper presents an efficient servomotor-aided calibration method for the triaxial gyroscope. The entire calibration process only requires approximately one minute, and does not require high-precision equipment. This method is based on the idea that the measurement of the gyroscope should be equal to the rotation speed of the servomotor. A six-observation experimental design is proposed to minimize the maximum variance of the estimated scale factors and biases. In addition, a fast converging recursive linear least square estimation method is presented to reduce computational complexity. The simulation results reflect the robustness of the calibration method under normal and extreme conditions. We experimentally demonstrate the feasibility of the proposed method on a robot arm, and implement the method on a microcontroller. We verify the calibration results of the proposed method by comparing with a traditional turntable approach, and the experiment indicates that the results of these two methods are comparable. By comparing the calibrated low-cost gyroscope reading with the reading from a high-precision gyroscope, we can conclude that our method significantly increases the gyroscope’s accuracy.

Ensemble Feature Selection for the Recognition of Human Activities and Postural Transitions on Smartphones

Human activity recognition (HAR) plays a vital role in the field of ambient assisted living for the welfare of senior citizens which provides satisfactory results in terms of wearable devices, ambient sensors, and smartphones. The usage of smartphones is familiar among users and it is widely adopted in activity recognition due to its low-cost, intrusiveness, and convenience. The embedded inertial sensors such as tri-axial accelerometer and gyroscopes are utilized to analyze the activities through the generated time and frequency domain features. Various state-of-the-art techniques have been proposed for significant feature selection to improve recognition performances. This paper proposes an ensemble of feature selection technique which selects minimal features from the intersection of features evolved in various other feature selection techniques. Extensive experimentation has been carried out using the HAPT dataset to identify the significant difference with state-of-the-art techniques through familiar classification performance metrics.

Complementary Filter for Attitude Estimation Based on MARG and Optical Flow Sensors

The combination of tri-axial magnetometer, accelerometer, and gyroscope has been widely used for three-dimensional attitude measurement, and this combination is also called MARG sensor. However, the measurement of accelerometer contains both the gravitational and motional acceleration, and only the former is useful for attitude estimation. As a result, MARG-based attitude estimation is easily disturbed by motion acceleration. In this paper, we introduce a complementary filter that estimates gravity and geomagnetic vectors in parallel, and utilize an optical flow sensor to detect and compensate motion acceleration. Experiment results show that the proposed algorithm has better performance than the existing ones when experiencing linear acceleration.

Precision of inertial measurement unit sensors in head-tracking systems used for binaural synthesis

Binaural synthesis is the most used sound system for diverse virtual and augmented reality systems nowadays, given its simplicity of implementation and the need of using only two audio channels. It is widely used in computer games, auralization and even audio production. To achieve the most natural sound field recreation, systems used for binaural synthesis must include a head-tracking sensor to dynamically calculate the binaural signal for the head orientation at any given moment. This is done by inertial measurement unit (IMU) sensors, specifically the triaxial accelerometers, gyroscopes, and magnetometers. Simpler systems, e.g., Arduino or other embedded systems, provide only raw sensor data, and the orientation is calculated by a processing unit. Other, more complex systems such as smartphones or VR headsets already calculate their position from the raw data using complex sensor integration algorithms. In this paper, a measurement procedure for measuring the precision of IMU sensors is presented. The need of absolute orientation calibration is addressed, and challenges of sensor data drift are discussed. Measurement results for simple embedded systems and complex systems found in smartphones are presented, and an estimation of IMU sensor quality for binaural synthesis is made.

Rapid Prototyping of Inertial MEMS Devices through Structural Optimization.

In this paper, we propose a novel design and optimization environment for inertial MEMS devices based on a computationally efficient schematization of the structure at the a device level. This allows us to obtain a flexible and efficient design optimization tool, particularly useful for rapid device prototyping. The presented design environment—feMEMSlite—handles the parametric generation of the structure geometry, the simulation of its dynamic behavior, and a gradient-based layout optimization. The methodology addresses the design of general inertial MEMS devices employing suspended proof masses, in which the focus is typically on the dynamics associated with the first vibration modes. In particular, the proposed design tool is tested on a triaxial beating-heart MEMS gyroscope, an industrially relevant and adequately complex example. The sensor layout is schematized by treating the proof masses as rigid bodies, discretizing flexural springs by Timoshenko beam finite elements, and accounting for electrostatic softening effects by additional negative spring constants. The MEMS device is then optimized according to two possible formulations of the optimization problem, including typical design requirements from the MEMS industry, with particular focus on the tuning of the structural eigenfrequencies and on the maximization of the response to external angular rates. The validity of the proposed approach is then assessed through a comparison with full FEM schematizations: rapidly prototyped layouts at the device level show a good performance when simulated with more complex models and therefore require only minor adjustments to accomplish the subsequent physical-level design.

Low-cost MIMU based AMS of highly dynamic fixed-wing UAV by maneuvering acceleration compensation and AMCF

Attitude Measurement System (AMS) that constructed by the low-cost Micro-Inertial Measurement Unit (MIMU) is usually adopted as the backup equipment for the fixed-wing Unmanned Aerial Vehicle (UAV) to enhance its safety performance during flight. However, both the large-maneuvering acceleration and the high-frequency propeller dithering are major reasons that decreasing its real-time measurement precision. In addition, there are only tri-axial gyroscope and tri-axial accelerometer measurements could be utilized to compensate the measurement errors in the autonomous MIMU based AMS. In this paper, an Adaptive Mahony Complementary Filter (AMCF) with dynamic tuning of the complementary filter Proportion and Integral (PI) parameters is used to estimate the real-time attitude of the highly dynamic aviation UAV with low-cost MIMU in high-frequency propeller dithering. Meanwhile, the maneuvering acceleration compensation technology is also integrated with the AMCF to eliminate the UAV highly dynamic maneuvering acceleration. Moreover, the attitude of AMS is updated by the quaternion updating algorithm to improve its real-time performance and reliability. Finally, both UAV ground high-speed taxiing experiment and UAV flight experiment are conducted to verify the measurement accuracy of low-cost AMS. The real-time flight experimental results demonstrated that the Rooted Mean Square Error (RMSE) of AMS based on the low-cost MIMU do not exceed 0.882° in terms of pitch angle and 0.864° in roll angle under various maneuvering conditions. These results provide powerful supports for both the UAV developers and the low-cost MIMU based AMS.

Deep learning-based cattle behaviour classification using joint time-frequency data representation

In this paper, a sequential deep neural network in conjunction with a joint time-frequency domain data representation is explored for the problem of cattle behaviour classification. The experimental evaluation is based on a real-world dataset with over 3 million samples, collected from sensors with a tri-axial accelerometer, magnetometer and gyroscope, attached to the collar tags of 10 beef steers. The experimental result demonstrate that the time–frequency domain data representation allows to efficiently trade-off a large reduction of model size and computational complexity for a very minor reduction in classification accuracy. This shows the potential of this classification approach to run on resource-constrained embedded and IoT devices. Most importantly, the proposed behaviour classification method achieves a high classification performance with an F1 Score of 94.9% for 3 behaviour classes, and 89.3% for 9 behaviour classes. This is in comparison to the current state-of-the-art with an F1 Score of 94.3% (for two classes) and 88.7% (for 8 classes).

Is a Head-Worn Inertial Sensor a Valid Tool to Monitor Swimming?

This study aimed to independently validate a wearable inertial sensor designed to monitor training and performance metrics in swimmers. A total of 4 male (21 [4]y, 1 national and 3 international) and 6 female (22 [3]y, 1 national and 5 international) swimmers completed 15 training sessions in an outdoor 50-m pool. Swimmers were fitted with a wearable device (TritonWear, 9-axis inertial measurement unit with triaxial accelerometer, gyroscope, and magnetometer), placed under the swim cap on top of the occipital protuberance. Video footage was captured for each session to establish criterion values. Absolute error, standardized effect, and Pearson correlation coefficient were used to determine the validity of the wearable device against video footage for total swim distance, total stroke count, mean stroke count, and mean velocity. A Fisher exact test was used to analyze the accuracy of stroke-type identification. Total swim distance was underestimated by the device relative to video analysis. Absolute error was consistently higher for total and mean stroke count, and mean velocity, relative to video analysis. Across all sessions, the device incorrectly detected total time spent in backstroke, breaststroke, butterfly, and freestyle by 51% (15%). The device did not detect time spent in drill. Intraclass correlation coefficient results demonstrated excellent intrarater reliability between repeated measures across all swimming metrics. The wearable device investigated in this study does not accurately measure distance, stroke count, and velocity swimming metrics or detect stroke type. Its use as a training monitoring tool in swimming is limited.

Improved Single Inertial-Sensor-Based Attitude Estimation during Walking Using Velocity-Aided Observation.

This paper presents a Kalman filter-based attitude estimation algorithm using a single body-mounted inertial sensor consisting of a triaxial accelerometer and triaxial gyroscope. The proposed algorithm has been developed for attitude estimation during dynamic conditions such as walking and running. Based on the repetitive properties of the velocity signal of human gait during walking, a novel velocity-aided observation is used as a measurement update for the filter. The performance has been evaluated in comparison to two standard Kalman filters with different measurement update methods and a smoother algorithm which is formulated in the form of a quadratic optimization problem. Whereas two standard Kalman filters give maximum 5 degrees in both pitch and roll error for short walking case, their performance gradually decrease with longer walking distance. The proposed algorithm shows the error of about 3 degrees in 15 m walking case, and indicate the robustness of the method with the same performance in 75 m trials. As far as the accuracy of the estimation is concerned, the proposed method achieves advantageous results due to its periodic error correction capability in both short and long walking cases at varying speeds. In addition, in terms of practicality and stability, with simple parameter settings and without the need of all-time data, the algorithm can achieve smoothing-algorithm-performance level with lower computational resources.

Assessment of Shoulder Range of Motion Using a Wearable Inertial Sensor Network

Diagnosis and progress monitoring of frozen shoulder rely on correct measurement of active and passive range of motion (ROM) of shoulder. This paper studied the feasibility of a wearable inertial sensor network (WISN) and its associated shoulder ROM estimation algorithm for shoulder ROM assessment. The WISN composed of three inertial modules were placed on the trunk, upper arm, and forearm of human for assessment of shoulder ROM in real time. Each inertial module consists of an ARM-based 32-bit microcontroller, a triaxial accelerometer, a triaxial gyroscope, a triaxial magnetometer, and a controller area network (CAN) transceiver. The measured accelerations, angular velocities, and magnetic signals generated by the human shoulder movements are transmitted to a personal computer via a Bluetooth wireless transmission module. The proposed shoulder ROM estimation algorithm includes the procedures of data collection, signal preprocessing, quaternion-based orientation estimation, and shoulder joint compensation. In order to evaluate shoulder ROM accurately, accelerations, angular velocities, and magnetic signals are integrated into a quaternion-based complementary nonlinear filter for minimizing the cumulative errors caused by the drift of the inertial sensors. Experimental results demonstrate that the WISN with the designed shoulder ROM estimation algorithm, compared to an optical motion analysis system (Vicon), have the average root mean square (RMS) angle errors ranging from 2.53° to 3.56°. In addition, the intra-tester reliability for each shoulder ROM measurement is excellent. The proposed WISN is a valid and reliable tool and can be used anywhere without any external reference device for convenient evaluation of shoulder ROM.

A smart point-of-care compliance monitoring solution for brace treatment of adolescent idiopathic scoliosis patients

Scoliosis is a medical condition which occurs in adolescents, where an individual's spine develops curvature. A Thoracolumbosacral orthosis (TLSO) is a type of brace used as a long term treat-ment method to control the lateral curvature of the spine in adolescent idiopathic scoliosis (AIS). However, compliance of the daily prescribed duration and tightness of the brace is a big challenge, which plays a significant role in the success of brace treatment. In this paper, we designed and de-veloped a wearable multi-modal sensor solution, which is embedded into the AIS patient's brace. The custom-designed hardware consists of a sensor board, a force sensor & housing, a miniature triaxial accelerometer and gyroscope sensor. The force sensor records the force being exerted by the patient's brace padding, while the accelerometer and gyroscope sensors capture upper trunk movements, providing cues to determine the patient's activities and lifestyle to help with more accurate estimation of the two key parameters, including 1) the in-brace tightness level 2) the duration of the brace wear, in an unsupervised manner. Furthermore, we presented a signal pro-cessing and data analytic methodology to identify the duration and tightness of the brace wear, in addition to identifying various patient activities based on the fusion of continuous force and iner-tial motion sensor recordings. We, specifically, derived a signal processing activity identification method using power spectral density estimation of the sensor data recordings. Then, we evaluate of the effectiveness of brace treatment in a pervasive manner. The proposed method evaluates the duration of brace wear through the process of segmentation and calculates the tightness level of the brace by estimating the baseline force per segment of data. This analysis is done when the patient performs activities including sitting, standing, climbing, walking, running and lying. We investigated an experimental scenario in which the patient performs a series of pre-defined activ-ities at home during day-long segments of brace wear. All analysis is performed using pervasive sensor data recordings. The experimental results demonstrated that we achieved an overall accu-racy of 99.8%, 100% and 99.9% for patients 1, 2 and 3 respectively for semi-supervised activity detection. We continued our compliance study for approximately a month. The level of tight-ness of brace-fit gradually reduced over a period of 2 weeks by 30% as the compliance of brace treatment increased from 7% to 90% on average. Given our experimental results and objective observations, our proposed system is capable of arranging for re-fitting sessions automatically so that the physician can adjust the brace tightness levels for a more effective brace treatment. • Brace treatment for AIS patients is long-term & pervasive demanding self-compliance. • Pervasive monitoring can provide two key factors of duration and tightness of the brace-wear. • We design a multimodal smart brace monitoring system to estimate the two factors. • We design a robust signal processing recipe to undertake unsupervised recordings. • Our experiments demonstrated that our system generates reliable & robust estimates.

Adaptive interval type-2 fuzzy recurrent RBFNN control design using ellipsoidal membership functions with application to MEMS gyroscope

In this article, a hybrid control approach is provided to control the micro-electro-mechanical system (MEMS) triaxial gyroscope as a multi-input multi-output (MIMO) system. Control design includes a fast non-singular terminal sliding mode control (FNTSMC) as a main part of the proposed hybrid control method and since the MEMS gyroscope performance is affected by parameter variations, quadrature errors, and external disturbances in the core of the main controller, adaptive interval type-2 recurrent fuzzy radial basis function neural network (IT2-RFRBF-NN) is employed to estimate the lumped uncertainties. The proposed hybrid approach has four main attributes: (1) it lies in the category of model-free control structures; (2) There is no negative power involved. Hence, the suggested method does not have the singularity problem; (3) to enhance the capability of the proposed method in the present of noise the ellipsoidal membership functions are employed to design adaptive IT2-RFRBF-NN; (4) the Fourier series expansion as a function approximation technique is efficiently used to online estimate the discontinuous component by establishing a soft switching in the proposed controller. With the help of the Lyapunov stability theory, guaranteeing the closed-loop control system stability by the suggested control design is confirmed. The findings of the simulations and comparison with other approaches confirm the superiority of the suggested hybrid approach.

Human Activity Recognition through Recurrent Neural Networks for Human–Robot Interaction in Agriculture

The present study deals with human awareness, which is a very important aspect of human–robot interaction. This feature is particularly essential in agricultural environments, owing to the information-rich setup that they provide. The objective of this investigation was to recognize human activities associated with an envisioned synergistic task. In order to attain this goal, a data collection field experiment was designed that derived data from twenty healthy participants using five wearable sensors (embedded with tri-axial accelerometers, gyroscopes, and magnetometers) attached to them. The above task involved several sub-activities, which were carried out by agricultural workers in real field conditions, concerning load lifting and carrying. Subsequently, the obtained signals from on-body sensors were processed for noise-removal purposes and fed into a Long Short-Term Memory neural network, which is widely used in deep learning for feature recognition in time-dependent data sequences. The proposed methodology demonstrated considerable efficacy in predicting the defined sub-activities with an average accuracy of 85.6%. Moreover, the trained model properly classified the defined sub-activities in a range of 74.1–90.4% for precision and 71.0–96.9% for recall. It can be inferred that the combination of all sensors can achieve the highest accuracy in human activity recognition, as concluded from a comparative analysis for each sensor’s impact on the model’s performance. These results confirm the applicability of the proposed methodology for human awareness purposes in agricultural environments, while the dataset was made publicly available for future research.

Continuous home monitoring of Parkinson's disease using inertial sensors: A systematic review.

Parkinson's disease (PD) is a progressive neurological disorder of the central nervous system that deteriorates motor functions, while it is also accompanied by a large diversity of non-motor symptoms such as cognitive impairment and mood changes, hallucinations, and sleep disturbance. Parkinsonism is evaluated during clinical examinations and appropriate medical treatments are directed towards alleviating symptoms. Tri-axial accelerometers, gyroscopes, and magnetometers could be adopted to support clinicians in the decision-making process by objectively quantifying the patient's condition. In this context, at-home data collections aim to capture motor function during daily living and unobstructedly assess the patients' status and the disease's symptoms for prolonged time periods. This review aims to collate existing literature on PD monitoring using inertial sensors while it focuses on papers with at least one free-living data capture unsupervised either directly or via videotapes. Twenty-four papers were selected at the end of the process: fourteen investigated gait impairments, eight of which focused on walking, three on turning, two on falls, and one on physical activity; ten articles on the other hand examined symptoms, including bradykinesia, tremor, dyskinesia, and motor state fluctuations in the on/off phenomenon. In summary, inertial sensors are capable of gathering data over a long period of time and have the potential to facilitate the monitoring of people with Parkinson's, providing relevant information about their motor status. Concerning gait impairments, kinematic parameters (such as duration of gait cycle, step length, and velocity) were typically used to discern PD from healthy subjects, whereas for symptoms' assessment, researchers were capable of achieving accuracies of over 90% in a free-living environment. Further investigations should be focused on the development of ad-hoc hardware and software capable of providing real-time feedback to clinicians and patients. In addition, features such as the wearability of the system and user comfort, set-up process, and instructions for use, need to be strongly considered in the development of wearable sensors for PD monitoring.