Search

The zero-velocity update (ZUPT) method is an effective technique, which has been widely used in low-cost foot-mounted inertial navigation systems. It can mitigate the accumulated error by using the fact that, when the foot is stationary, the velocity should be zero. However, it is hard to detect the zero-velocity phases (ZVPs) correctly in various motion modes. To deal with this problem, a novel adaptive zero velocity detection algorithm based on an improved general likelihood ratio test (IGLRT) detector is proposed. The prior knowledge of the proposed IGLRT detector is the minima output value of inertial sensors, which is determined by an improved peak-valley detection method, while the prior knowledge of the traditional general likelihood ratio test (GLRT) method is based on the hypothesis that, when the foot is stationary, the magnitude of angular rates is zero, and the magnitude of accelerometers is equal to the magnitude of gravitation. Also, an adaptive method is proposed to enhance the robustness of the IGLRT, where the thresholds of IGLRT can be adjusted adaptively according to the amplitude of accelerometers. Field tests show that the proposed algorithm performs better than conventional and state-of-the-art approaches in mixed motion modes. [ABSTRACT FROM AUTHOR]

In this work, an optimized tunnel magnetoresistive (TMR) accelerometer with a closed-loop control system was developed and evaluated. The device first introduces a silicon spring–mass sensing structure lower than 50 Hz into TMR-based accelerometry for enhancing the mechanical sensitivity and subsequent readout sensitivity. Simultaneously, in order to realize the in-plane electrostatic feedback control, the comb structure is designed along with the sensing mechanism, owning combined benefits of integrated processing and large feedback force. The whole sensing structure is a silicon-glass chip, fabricated by the standard micro-electromechanical system (MEMS) process—deep dry silicon on glass (DDSOG) process. A permanent rubber magnet is assembled on the proof mass for conversion from the displacement to variation of the magnetic field intensity, which is further detected by a pair of symmetrically arranged TMR sensors. The voltage signals output from TMR sensors are then sent into an analog circuit via an interface module for force-feedback control. The simulation analysis indicates that the proposed MEMS sensing structure has a low natural frequency of 44.55 Hz, corresponding to a compliant mechanical sensitivity of 125.5 $\mu \text{m}$ /g. Meanwhile, a maximum magnetic sensitivity of about 0.1 mT/mm is available in a height of 6 mm above the $3\times 3\times0.3$ mm magnet. Finally, the experiments on the assembled prototype demonstrated that a scale factor of 1.79 V/g and a bias stability of 228 $\mu \text{g}$ have been achieved in the closed-loop modality, which verifies the effectiveness of the proposed TMR MEMS accelerometer. [ABSTRACT FROM AUTHOR]

A popular approach to track the position of a user is to use an inertial measurement unit (IMU), which allows tracking a user with a pedestrian dead reckoning (PDR) system by estimating the length and heading of each step a user takes. Furthermore, a user’s steps together with their length can be used to detect anomalies in the gait due to, e.g., a gait disorder. To determine the length and/or heading of a step, we first need an algorithm that estimates the boundaries of a step, i.e., the start and end, in the IMU signals. In this article, we propose a deep-learning-based step detection algorithm that detects the start and end of a step using only data from a handheld accelerometer. In this algorithm, also, the relationship between the start and end of a step is used to improve performance. To evaluate the performance of the detector, in contrast to most previous works that only look at the total number of detected steps, we determine if the start and end of each step are detected at the correct instant. This resulted in a step detector with an f-score of 99.2% and 99.0% for, respectively, detecting the start and end of a step. After detecting the start and end of the steps, we use a long short-term memory (LSTM)/convolutional neural network (CNN)-based step length estimator, which resulted in a mean absolute error (mae) on the step length of 3.21 cm. Finally, we also determine the combined performance of the step detector and step length estimator, i.e., where the proposed step detector is used to extract the accelerometer fragments of a step instead of assuming that the boundaries are known. This resulted in an mae of 6.56 cm using the deep-learning-based step length estimator. [ABSTRACT FROM AUTHOR]

Due to the COVID-19 global pandemic, there is more need for remote patient care, especially in rehabilitation requiring direct contact. However, traditional Chinese rehabilitation technologies, such as gua sha, often need to be implemented by well-trained professionals. To automate and professionalize gua sha, it is necessary to record the nursing and rehabilitation process and reproduce the process in developing smart gua sha equipment. This article proposes a new signal processing and sensor fusion method for developing a piece of smart gua sha equipment. A novel stabilized numerical integration method based on information fusion and detrended fluctuation analysis (SNIF-DFA) is performed to obtain the velocity and displacement information during the gua sha operation. The experimental results show that the proposed method outperforms the traditional numerical integration method with respect to information accuracy and realizes accurate position calculations. This is of great significance in developing robots or automated machines that reproduce the nursing and rehabilitation operations of medical professionals. [ABSTRACT FROM AUTHOR]

An in-fiber Fabry–Perot interferometer (FPI) is proposed and designed for performance improvement of a fiber Bragg grating (FBG) accelerometer. It consists of two FBGs which are inscribed using the femtosecond laser point-by-point technology. The spectral characteristics of the FBG Fabry–Perot (FBG-FP) were studied. Edge filter interrogation configuration was employed in the vibration measurement system. In contrast to the conventional FBG accelerometer used in wavelength interrogation or edge filter interrogation configurations, the scheme has a higher sensitivity and lower detection threshold which are achieved improving the optical properties of the accelerometer. We fabricated a high-frequency accelerometer and experiments were carried out to verify its performance. The sensitivity and detection threshold were 27.3 mV/g and 1.5 mg, respectively. This study presents a feasibility analysis of the short cavity FBG-FP, as opposed to a single FBG, for the vibration accelerometer. [ABSTRACT FROM AUTHOR]

This article presents a noninvasive method of classifying gait patterns associated with various movement disorders and/or neurological conditions, utilizing unobtrusive, instrumented socks and a deep-learning network. Seamless instrumented socks were fabricated using three accelerometer-embedded yarns, positioned at the toe (hallux), above the heel, and on the lateral malleolus. Human trials were conducted on 12 able-bodied participants, an instrumented sock was worn on each foot. Participants were asked to complete seven trials consisting of their typical gait and six different gait types that mimicked the typical movement patterns associated with various movement disorders and neurological conditions. Four neural networks and an SVM were tested to ascertain the most effective method of automatic data classification. The bi-long short-term memory (LSTM) generated the most accurate results and illustrates that the use of three accelerometers per foot increased classification accuracy compared to a single accelerometer per foot by 11.4%. When only a single accelerometer was utilized for classification, the ankle accelerometer generated the most accurate results in comparison to the other two. The network was able to correctly classify five different gait types: stomp (100%), shuffle (66.8%), diplegic (66.6%), hemiplegic (66.6%), and “normal walking” (58.0%). The network was incapable of correctly differentiating foot slap (21.2%) and steppage gait (4.8%). This work demonstrates that instrumented textile socks incorporating three accelerometer yarns were capable of generating sufficient data to allow a neural network to distinguish between specific gait patterns. This may enable clinicians and therapists to remotely classify gait alterations and observe changes in gait during rehabilitation. [ABSTRACT FROM AUTHOR]

Magnetic and inertial measurement units (MIMUs) are mainly used for determining the attitude of moving bodies. The extended Kalman filter (EKF) and the complementary filters (CFs) are the most commonly applied algorithms for calculating the attitude of the body that an MIMU is attached. Despite its much higher accuracy than CFs, the time-consuming EKF may not be competent for the work of attitude calculation if a microcomputer with limited computation power is used. To any attitude calculating algorithm, its posterior attitude (PA) is computed as: PA = prior attitude + gain matrix (GM) $\times $ innovation. The prior attitude and the innovation are derived from the measurements of the gyroscope, the accelerometer, and the magnetometer. The detailed analysis in this study found that the reason for the performance difference between EKF and CFs is their different structures of the GMs, which promotes the authors to construct a new GM. Based on some reasonable assumptions, the GM of EKF is simplified under the principle that the computation accuracy of PA is decreased as little as possible, but its computation burden is reduced as much as possible. After the replacement of the GM of the EKF by the innovatively designed GM, a new CF, named time-efficient complementary Kalman gain filter (TCF), is proposed. The simulation and experimental results show that TCF performs better than two CFs and is close to EKF in terms of attitude estimation accuracy but uses nearly the same computation time as the two CFs and about half of the computation time required by EKF. [ABSTRACT FROM AUTHOR]

As civil structures are exposed to various external loads, their periodic evaluation is paramount to ensure their safety. By estimating the 6-degree-of-freedom (DOF) displacement of structures, structural behavior can be monitored directly. Therefore, this study aims to develop a translational and rotational displacement estimation method by fusing a vision sensor and inertial measurement unit (IMU) using a quaternion-based iterative extended Kalman filter (QIEKF). The QIEKF algorithm was applied to reduce the nonlinear influence on the measurement model. The 6-DOF displacement is predicted using the integral of the gyroscope output and updated via a combination of an accelerometer and a magnetometer through a vector matching process in the Kalman filter framework. Subsequently, the 6-DOF displacement estimation result is updated through a vision sensor using a 2-D planar marker and homography transformation in the Kalman filter framework. The performance of the proposed sensor fusion method was verified with experiments using a motorized motion stage, and the results show that the displacements can be estimated with high accuracy regardless of measurement noise and slowly varying signal drift. [ABSTRACT FROM AUTHOR]

Multiplexing technology is important when realizing multipoint measurement. We propose a vector accelerometer based on a silicone rubber compliant cylinder, which makes it easy to realize multistage cascaded multiplexing. The proposed accelerometer has the natural frequencies of 19–39 Hz, which increase with an increased Shore A material hardness and a shortened length of the intermediate part of the compliant cylinder. In addition, at a random orientation, the maximum sensitivity reaches 340.6 pm/g when subjected to a 10-Hz vibration excitation. Moreover, the accelerometer can recognize the vibration orientation from 0° to 360°. With high integration, small size, and multistage cascaded multiplexing, the accelerometer is, especially, suitable for narrow spaces multipoint measurement. [ABSTRACT FROM AUTHOR]

This article proposes a single-axis, highperformance Complementary Metal Oxide Semiconductor- Microelectromechanical System (CMOS-MEMS) closed-loop accelerometer, implemented using in-plane movable suspended gate-field-effect transistor (IP-SGFET). The IPSGFET’s gate is embedded within a movable proof mass supported by flexible folded beams. The external acceleration displaces the proofmass, which changes the drain current of IP-SGFET. The change in current is read out with the help of a simple differential amplifier which is constructed with the IP-SGFET. The differential amplifier’s output voltage is used througha closed-loopintegral controland a high-voltage dynamically reconfigurable charge pump, which generates a high-voltage differential drive of up to 10.5 ± 5.12 V, to drive an actuator that brings the gate back to its resting position. The actuator used is a comb-drive type electromechanical actuator. In this work, two tools, viz., MEMS+ and Technology Computer Aided Design (TCAD) device simulation tools are used in a sequential manner to handle the electrostatics coupled with the micro-mechanical domain present in the suspended gate field effect transistors (SGFET). The results obtained are then used as look-up-tables for the design and simulation of the entire system, including the control circuits, in SPECTRE. Appropriate design of the MEMS suspended structure, the field effect transistor (FET) device, and interface circuit results in a open-loop sensitivity of 927 mV/g with a near-zero cross-axis sensitivity. Using the actuator and the reconfigurable charge pump, closed-loop operation is realized with a dynamic range of ±4 g at a supply voltage of 3.3 V. [ABSTRACT FROM AUTHOR]

For high-impact devices, subsequent vibrations have as much influence on the deterioration of the mechanical structure as the impact itself. To mitigate the consequences of both impacts and resulting vibrations, it is crucial to accurately understand the peak acceleration and vibration frequencies that originate in the structure. In this article, a radiation tolerant opto-mechanical sensor based on an absolute measurement inertial accelerometer is presented. This method of measurement is chosen due to the viability of placing the readout electronics far from the highly radioactive environment that the sensor may be installed in. The designed accelerometer consists of a fixed aluminum frame with a built-in membrane that acts as a leaf spring. In the same membrane, a retro-reflector is attached which acts as a seismic mass. The retro-reflector reflects the light beam coming from the collimator lens placed on the frame. To prove its viability, the device has been tested under different conditions. First, the device is calibrated and validated using a high-frequency exciter. Second, a high-acceleration testbench is used to compare its output signal with a commercial accelerometer and verify its proper operation as well as its expected lifetime. The experimental results showed an average sensitivity of ${9} \times {10}^{-{4}}$ g/pm with reading measurements of up to 5000 g and a working bandwidth located between 50 and 800 Hz. [ABSTRACT FROM AUTHOR]

Directional drilling utilizes the measurement-while-drilling (MWD) to provide attitude for drilling bit. A traditional MWD system, consisting of magnetometers and accelerometers, relies on expensive and huge nonmagnetic drill collar (NMDC) to shield from magnetic interference to obtain accurate attitude angles. A fiber optic gyroscope-based MWD (FOG-based MWD) that is based on inertial navigation is an alternative survey scheme. But, random vibration environment of underground introduces high-amplitude random noise to the FOGs. Besides, the auxiliary measuring equipment [such as global position system (GPS)] will not be available in underground environments, which will bring massive attitude error under pure inertial navigation. Therefore, combined with the engineering application conditions, this article proposes a modified attitude precision improvement approach of the FOG-based MWD. First, an autoregressive moving average (ARMA) error model for FOG is established to restrain the random vibration environment interference of underground. Second, combined with the FOG error model, a continuous backtracking navigation error model is developed to improve the azimuth accuracy. Moreover, the continuous backtracking algorithm frame is designed to estimate the attitude. Finally, a semi-physical experiment and a drilling test verify that the error model has an obvious effect on reducing the amplitude of FOG random noise. The proposed backtracking algorithm make the azimuth angle achieve a high-level estimation in short time. The result of the drilling test shows that the azimuth has an improvement of 48.79%, as compared with the traditional zero velocity update (ZUPT). [ABSTRACT FROM AUTHOR]

A fast bowler in the game of cricket is exposed to high-intensity actions and is more prone to lumbar stress fractures, ankle sprains, and knee, shoulder, and lower back injuries. There are different phases of bowling, from running to ball release, viz., run-up, predelivery stride (PDS), mid-bound (MB), front foot contact (FFC), ball release, and follow through. The front leg motion of the bowler plays a crucial role in delivering the ball efficiently. An improper landing after the jump causes most of the injuries as more strain is induced on the knee, ankle, and groin region. Loss of balance during the follow-through is the other most common reason for injury. A braced front leg at FFC helps the bowler transfer the momentum gained during the run-up into a fast-bowling action. For a refined understanding of the motion of the bowler and the causes of injuries, the identification of bowling phases is crucial. This article proposes the design of a miniature and economical wearable sensor unit using a Bosch BNO055 inertial measurement unit (IMU) sensor that the players can comfortably wear during the game. A novel method for calculating knee flexion of bowlers at FFC using quaternion data gathered from sensors is presented. Two wearable units were placed on the bowler’s front leg to collect motion data. Supervised machine-learning algorithms have been evaluated for classifying bowling phases using statistical features extracted from the sensor data to assist coaches and players with data-driven and evidence-based bowlers’ training. [ABSTRACT FROM AUTHOR]

An inertial measurement unit (IMU) architecture is proposed by combining a set of accelerometers with a three-axis gyroscope, in closed-loop operation, to achieve superior accuracy and low noise estimation of the angular velocity by compensating for accelerometers’ bias and wandering. It advances the concept of gyroscope-free IMUs, which derive the angular velocity using multiple accelerometers in an open-loop configuration, by eliminating their constant drift errors due to accelerometers’ bias, enabling real-world applications. The stability of the proposed closed-loop system is studied analytically with sufficient conditions provided, and a closed-form expression of the angular velocity noise is derived. Both experimental and simulation results indicate that the proposed architecture excels in terms of accuracy and noise performance. [ABSTRACT FROM AUTHOR]

This study proposed a wearable inertial motion capture (Mocap) method for Parkinson’s lower limb motion reconstruction and analysis. To eliminate the error accumulation caused by sensor noise, the error state Kalman filter is used for attitude estimation. The inertial Mocap method proposed in this study is compared with the OptiTrack system to verify the motion tracking effect. The accuracy of different phase-detection methods is compared, and the highest recognition accuracy is 99.11% using the long short-term memory (LSTM) method. In addition, this study collected subjects in healthy control (HC), Parkinson’s disease (PD), and freezing of gait (FOG) groups for comparative analysis. The experimental results revealed differences in motion characteristics and kinematics between PD and healthy subjects. Thus, the quantitative analysis results can be used as an important reference for the clinical application of PD. [ABSTRACT FROM AUTHOR]

This article proposes a novel hinge-type fiber Bragg grating (FBG) accelerometer for measuring medium-frequency vertical vibrations that employ a single-notch circular flexure hinge as the elastomer of the accelerometer. The proposed elastomer allows for the simultaneous realization of high sensitivity, a compact structure, and a relatively wide working frequency range, all of which are critical for accelerometer performance. The working principle and detailed theoretical analysis of the FBG accelerometer are illustrated and conducted. Moreover, the sensitivity of the accelerometer is optimized further through a genetic algorithm in order to strike a balance between miniaturization and resonance frequency. Finite element analysis and experimental tests were carried out to validate the performance of the accelerometer. The results indicate that the proposed FBG accelerometer has a resonant frequency of 800 Hz and a working frequency range of 50–450 Hz, with a sensitivity of approximately 65 pm/g at an exciting frequency of 100 Hz between the fiber central wavelength and vibration acceleration. The results also demonstrate that the proposed accelerometer exhibits a good linear relationship with the frequency of vibration accelerations, as well as good anti-interference properties with the horizontal vibrations. It has the potential to provide a novel and simple solution for monitoring medium-frequency vibration signals in complex environments characterized by high levels of electromagnetic interference (EMI), such as in aeronautical hydraulic systems. [ABSTRACT FROM AUTHOR]

Antioverload high-range accelerometer is playing an increasingly important role in the field of hard target penetration testing, and the demand of higher accuracy and higher range accelerometers is growing consequently. Therefore, a novel accelerometer with a new structure of membrane island is proposed in this article, while the measurement range is significantly improved to 200 000 g. The design is mainly based on the finite-element method. In general, the membrane-island structured piezoresistive accelerometer is formed of a four-clamped membrane that carries a piezoresistor at the four ends. The four piezoresistors form a Wheatstone bridge. Then, based on the Timoshenko plate and shell theory, a mathematical stiffness model of the four-clamped membrane is established. The detailed parameters of the sensor are carried out by the finite-element analysis (FEA) and optimization, and it has a maximum stress of 26.8 MPa and a maximum displacement of $0.208~\mu \text{m}$ when applying an input acceleration of 200 000 g. Finally, the sensitivity test, antioverload test, and dynamic test of three membrane-island structured prototypes are carried out, and their sensitivities are 0.579, 0.505, and $0.532~\mu \text{V}$ /g, respectively; the passband with 5% dynamic error is 14.9, 16.5, and 15.8 kHz, respectively. The accelerometer has an antioverload capability of 250 000 g, which is verified by the Hopkinson bar. The test results show that this membrane-island structured accelerometer has both good antioverload capability and high accuracy. [ABSTRACT FROM AUTHOR]

In order to meet the needs of accelerometers in vibration monitoring, a fiber Bragg grating (FBG) accelerometer based on a miniature slide rail is proposed. The accelerometer consists of two FBGs, a miniature slide rail, two spring plungers, and a mass block. Since the motion of the two fiber gratings is in opposite directions, the induced displacement of the center wavelength can wipe one another out. So, the effect of temperature can be reduced and the sensitivity of the fiber grating can be improved. Through theoretical analysis and experimental verification, the resonant frequency, sensitivity, and lateral immunity of the sensor are studied. The experimental results show that the effect of temperature on the sensor structure is reduced to 0.3 pm/°C in the range of 22.5 °C–90 °C. The resonant frequency of the accelerometer is 200 Hz, and the FBG accelerometer has a broad frequency range from 0.5 to 150 Hz, with a corresponding sensitivity range from 307.3 to 319.2 pm/g, and the lateral immunity is less than 0.5%. The sensor is simple in structure and easy to manufacture. Therefore, it has great prospects for application in structural health monitoring and other fields. [ABSTRACT FROM AUTHOR]

An adaptive orientation estimation method based on hidden Markov model (HMM) using multiple Magnetic Angular Rate Gravity (MARG) sensor is proposed to address the problem of linear acceleration and ferromagnetic disturbance. Two parallel Kalman filters containing first-order Gauss-Markov disturbance models are implemented to estimate the disturbances before estimating the orientation. It is important that the noise covariance of the disturbance model should match the magnitude of the disturbance to ensure the estimation performance; otherwise, the disturbance estimation accuracy will reduce significantly. Therefore, an adaptive Kalman filter based on HMM is designed to adjust the disturbance noise covariance. After the disturbances are separated, the estimated gravity acceleration and geomagnetic field as well as the gyroscope measurement are taken as the inputs, and the orientation estimation procedure is established by using the multiplicative extended Kalman filter (MEKF). Experiments show that the proposed method achieves a higher level of accuracy in comparison with other orientation estimation methods under various conditions. [ABSTRACT FROM AUTHOR]

The estimation of the orientation of an object, and a human head in particular, can be defined by the Euler angles: the yaw, pitch and roll. The robust and drift-free estimation of those angles is usually achieved with the data from several sensors such as accelerometers, gyroscopes and magnetometers, processed with sensor fusion algorithms. However, wearable devices such as hearing instruments are rarely equipped with all those sensors, and might usually only have a single accelerometer embedded per device. While it is possible to retrieve a correct estimation of the roll and pitch using only a single accelerometer, estimating the yaw is a more challenging task as accelerations around the gravity vector cannot be detected by an accelerometer. In the context of binaural communication devices and spatial hearing, the yaw of the head is a key information that helps achieving dynamic binaural synthesis. This work proposes an algorithm that aims at estimating the yaw of a human head by using only two 3-axis accelerometers that are placed at each side of the head. The algorithm is evaluated with acceleration measurements of human subjects achieving a realistic task. The results suggest that the strategy used is promising as a good performance can be achieved for a majority of the measures as long as an individualized set of parameters can be selected. The results also suggest that the performance of the current implementation is sensitive to sensor displacement after calibration or a wrong estimation during initialization. [ABSTRACT FROM AUTHOR]

The integrated strap-down inertial/celestial navigation system (SINS/CNS) has been widely applied in autonomous navigation. For long-term high precision performance, initialization is a vital process including the calibration of misalignment angles and system-level parameters. This paper proposes a decoupling three-position calibration method based on a Kalman filter. By measuring the velocity error and attitude error, it is capable to estimate the attitude angle, gyro bias, accelerometer bias and installation error of star sensor. Firstly, the observability degree analysis provides a theoretical account of the coupling of states and the effect of different rotation maneuvers. Secondly, the limitation of gyro noise on the accuracy of accelerometer bias is clarified by comparative experiments. Simultaneously, results indicate that the proposed method can achieve arc-second accuracy in attitude angle and installation error angle, and limit the accelerometer bias error to less than ${3}\%$ regardless of the IMU performance. It is an optimal calibration method with advantages of fast convergence, high accuracy, good stability and repeatability. [ABSTRACT FROM AUTHOR]

Low cost and small size integrated sensors bring a significant interest to redundant sensing systems such as sensor array systems. Redundancy may permit to increase both performance and dependability of individual sensors at the system level, however, the emergence of these redundant systems has generated the need for dynamic algorithms able to detect and act against the presence of errors in system elements. Then, a new adaptive algorithm for sensor array systems is presented in this work. This approach is generic and can be implemented for different types of sensor systems, besides it does not require any information about the inputs. The proposed real-time algorithm is based on the following: 1) estimation of sensor individual variances at each time step, 2) identification and removal of sensors affected by errors, and 3) reincorporation of sensors after recovery or replacement. Variance estimation is revisiting MInimum Norm Quadratic Unbiased Estimation (MINQUE) method to allow real-time operation and requires that the number of sensors is strictly greater than two times the number of signals to be measured. Consequently, the proposed method is able to support the presence of m-1 faulty devices, where m is the total number of sensors less two times the number of measured signals. Performance of the proposed algorithm is demonstrated through theoretical demonstrations, simulations, and experimental results. [ABSTRACT FROM AUTHOR]

With the technological development in healthcare, environments such as rehabilitation clinics and patients’ houses, are increasingly monitored by multi-device systems. To aggregate information, overcome privacy issues and device failures, it is essential to match measurements from different sources and associate them to a particular patient. While cameras are used to detect and track anonymized persons, wearable devices can acquire inertial and health information. The challenge is to correctly pair the tracked persons to their status, such as heart rate, collected by other devices. Recently, many works have been proposed, in several scenarios, to tackle sensor fusion-based tracking using a large variety of information. However, when the budget is limited, the involved sensing devices lack of inertial components, such as gyroscope, and may have low precision. In this work, we propose a novel solution to match unlabeled 3D skeletons, detected by a depth camera, with on-wrist wearable devices equipped only with accelerometer. Additionally, a Deep Learning submodule, named SkeletonRNN, is introduced to overcome camera failures in the 3D skeletons points detection and fill missing joints. A complete dataset containing skeletons and accelerations measurements, of dailies and rehabilitation activities, has been collected, manually annotated, and is available for testing purposes. We trained and tested the SkeletonRNN using data augmentation on our dataset, the final average 3D point prediction error is 11.00cm and the skeleton-device pairing accuracy of the overall system is 76.62% on a total of 231 chucks. Datasets, code and experiments can be found at https://github.com/matteo-bastico/SkeletonRNN. [ABSTRACT FROM AUTHOR]

In this paper, a differential micro-opto-electro-mechanical system (MOEMS) accelerometer based on the Fabry-Pérot (FP) micro-cavities is presented. The optical system of the device consists of two FP cavities and the mechanical system is composed of a proof mass that is suspended by four springs. The applied acceleration tends to move the PM from its resting position. This mechanical displacement can be measured by the FP interferometer formed between the proof mass cross-section and the optical fiber end-face. The proposed sensor is fabricated on a silicon-on-insulator wafer using the bulk micromachining method. The results of the sensor characterization show that the accelerometer has a linear response in the range of $\boldsymbol {\pm }1\text{g}$. Also, the optical sensitivity and resolution of the sensor in the static characterization are 6.52 nm/g and $153 \boldsymbol {\mu }\text{g}$. The sensor sensitivity in the power measurement is 49.6 mV/g and its resonant frequency is at 1372 Hz. Using the differential measurement method increases the sensitivity of the accelerometer. Based on the experimental data, the optical sensitivity in static mode is two times as high as that of a similar MOEMS accelerometer with one FP cavity. [ABSTRACT FROM AUTHOR]

This paper develops an improved system-level calibration method for the strapdown inertial navigation system (SINS). It solves the problems of too many dimensions of state variables and installation error matrix coupling. To solve the two problems, the improved system-level calibration method derives the physical causes of the installation error matrix. The installation error matrix is decomposed into skew symmetric matrix and symmetric matrix. Then the appropriate system-level calibration scheme is designed based on the velocity error equations and attitude error equations. The designed system-level calibration scheme separates the coupled installation error matrix parameters by changing the value of strapdown matrix. Simulation results demonstrate that the proposed method is feasible. The relative errors of gyro calibration parameters are between 0.03% and 0.70%, and the relative errors of accelerometer calibration parameters are between 0.01% and 0.24%. For comparison, the navigation and positioning experiments are also performed to evaluate the proposed method. The results indicate that the improved system-level calibration scheme is better than the traditional system-level calibration scheme. The positioning error is reduced by more than 30%. [ABSTRACT FROM AUTHOR]

This paper describes a simplified analytical model of a damping constant ${b}$ to design a MEMS capacitive accelerometer using the multi-layer metal technology. The proposed model is introduced by combining the theoretical equation with the approximate form factor obtained by the measured data. In order to create the model, we fabricated several types of MEMS capacitive accelerometers with different structure parameters such as the etching hole area, the perforated proof-mass area, and the gap. The calculation results show that by the proposed model, the damping constant ${b}$ was in accord with the measured ${b}$. We also confirmed that the relative error between the measured ${b}$ and the damping constant ${b}$ by the proposed model could be improved to one-half of the conventional model. Moreover, the Brownian noise ${B}_{\text {N}}$ calculated by the proposed model was also consistent with the measured ${B}_{\text {N}}$. In addition, to confirm the suitability of the proposed model for CMOS-MEMS multi-physics simulation, we performed the simulation of the ring-down characteristics of a gold perforated proof-mass differential MEMS capacitive accelerometer. The simulated results suggested that the ring-down characteristics by the proposed model coincided with those of the measured data. Therefore, we verified that the proposed model would be effective for the analysis and the design of the MEMS capacitive accelerometer with the gold perforated proof-mass. [ABSTRACT FROM AUTHOR]

In this paper we present a state estimation scheme for Unmanned Aircrafts (UAs) utilizing dynamics based models and multi-sensor data fusion. Employing the UA dynamics in estimation can substantially enhance the estimator performance, but obtaining accurate dynamics parameters for each UA is computationally costly and complex. To eliminate these issues, we propose two decoupled Extended Kalman Filters (EKFs), namely the Rotational Decoupled Extended Kalman Filter (RDEKF) and the Translational Decoupled Extended Kalman Filter (TDEKF). The dynamics parameters in these filters are identified in real-time using the Deep Neural Network and the Modified Relay Feedback Test (DNN-MRFT) approach. This approach doesn’t demand prior knowledge of the UA physical parameters, requiring only an Inertial Measurement Unit (IMU) and a positioning system for model classification. Our estimation scheme provides position, velocity and attitude estimates, in addition to smooth lag-free inertial acceleration estimates. We show experimentally the advantages of our approach on trajectory tracking problems that uses low rate position sensors. We also demonstrate how utilizing the estimated acceleration in feedback control can reduce the tracking error of an optimally tuned system by 43%. Moreover, the proposed estimator produces smooth estimates that leads to a reduction of controller action by 6.6%, when compared to kinematic based estimators. We compare the achieved results against other methods that require full prior knowledge of the UA parameters or the noise models, and show advantages in performance and real-time capability. [ABSTRACT FROM AUTHOR]

Optical sensing techniques are well-known for electromagnetic immunity and thus have been widely considered in high-performance vibration sensors such as accelerometers and seismometers. In the progress of miniaturization, however, the force feedback control has rarely been developed due to associated complexity of the optical loops. Aiming to further explore the application of interferometry in geophysics, a force-rebalanced micro-optics accelerometer is developed in this work. This device operates relying on a grating-based MEMS interferometer, where a 3D sensing structure is microassembled to serve as the moveable reflector. In the meantime, a soft robber magnet is arranged on the proof mass and then the feedback mechanism is constructed in combination with a wound coil. The open-loop test indicates this feedback system can generate a drive force of 43.39 nN/mA, and possesses a resultant displacement-regulating capacity of 20.7 nm/mA. Implemented with a digital PI control, the closed-loop device demonstrates an improved dynamic range of 1.012 mg, equal to 10 times of the open-loop state, along with a scale factor of 3250 V/g and a $\text{R}^{{2}}$ coefficient of 0.99927. Additionally, the instrumental noise is tested as 10 ng/ $\sqrt {\text {Hz}}$ from 0.2 Hz to 10 Hz, and a bias stability of 36 ng is available at room temperature. Further considering the compact footprint of 4cm $\times $ 6cm $\times $ 3.15cm, this device exhibits great utility for microseismic observations in confined environments. [ABSTRACT FROM AUTHOR]

Rotary inertial navigation system(RINS) can suppress constant and slowly varying errors of inertial sensors through rotary modulation technology. However, the angle random walk(ARW) error of gyro can not be modulated, which has become the key factor of restricting the long-term navigation accuracy of RINS. In this paper, the influence of ARW on long-term navigation accuracy of RINS is analyzed. In view of the situation that GPS and other external auxiliary information are not available, an ARW suppression and error estimation and compensation method based on online weight distribution of multi-RINSs is proposed. Considering the disturbance of deflection to attitude and the disturbance of lever arm effect to velocity, the dynamic weighted average of position information of multi-RINSs are selected as the measurement, the equivalent error parameters of each RINS are estimated and compensated by KF filter is proposed. The simulation results show that the navigation errors caused by ARW can be reduced several times by using the proposed suppression method, and the weighted average position information to estimate and compensate the equivalent error of each RINS, which can further improve the navigation accuracy of each RINS by about 3 times, and proved that the proposed method of restraining ARW and error estimation and compensation is effective. It makes the relationship among the RINSs of from backup to auxiliary, which has significant value for further improving the autonomous long-term navigation accuracy. [ABSTRACT FROM AUTHOR]

The performance of absolute atom gravimeters used on moving platforms, such as vehicles, ships and aircrafts, is strongly affected by the vibration noise. To suppress its influence, we introduce a vibration compensation method, in which a classical accelerometer is used to measure the vibration noise. The measurement results obtained by the accelerometer show the vibration noise in the vehicle can be 2 orders of magnitude greater than that in the lab during daytime, and can induce an interferometric phase fluctuation with a standard deviation of $16.70\pi $. With the compensation method, our vehicle-mounted atom gravimeter can work normally under harsh conditions. Comparing the Allan standard deviations before and after the vibration noise correction, we find a suppression factor of 22.74 can be achieved in static condition with a half interrogation time of T = 20 ms, resulting in a sensitivity of 1.35 mGal/Hz1/2, and a standard deviation of 0.5 mGal with an average time of 10 s. We also demonstrate the first test of an atom gravimeter in a moving vehicle, in which a suppression factor of 50.85 and a sensitivity of 60.88 mGal/Hz1/2 were realized with T = 5 ms. [ABSTRACT FROM AUTHOR]

Wearable sensing has enabled physiological monitoring in everyday life. Clinically, cardiovascular and respiratory monitoring commonly requires separate devices. However, for feasibility in wearable systems, they are integrated into a single device. Since seismocardiography (SCG) records the motion of the chest wall, it has the potential to monitor both vital signs. The primary application of SCG is cardiovascular monitoring. As a result, its ability to detect respiration has been less defined. This study characterizes how SCG is affected by respiration and consequently, how respiratory metrics can be derived from these effects. The sternal motion of 40 subjects was recorded using an inertial measurement unit with a spirometer for reference. Three scenarios were examined with the subject at rest, holding their breath at peak inhalation, and holding their breath when exhaled. Three main aspects of respiratory modulation were explored: amplitude modulation, frequency modulation, and baseline wandering. The SCG amplitude was observed to be dependent on respiratory volume whereas baseline and frequency modulation were dependent on inhalation phase. All three effects were employed separately in the calculation of respiratory rate. Estimations from baseline wandering, amplitude modulation, and heart rate modulation produced r-squared values of 0.71, 0.44, and 0.66, respectively. The accuracy of tidal volume measurement was limited by a high inter-subject variability. Estimation of the respiratory waveform produced an average r-squared of 0.76 using multivariate linear regression. This demonstrates the potential of SCG as a tool for respiratory monitoring within an integrated, non-invasive cardiorespiratory monitoring system from a single point of contact. [ABSTRACT FROM AUTHOR]

The integrated navigation system of strapdown inertial navigation system (SINS) and odometer (OD) is a commonly used vehicle positioning system. The installation errors caused by the difference between the installation positions of SINS and OD mainly include the installation deflection angle and lever arm, which will reduce the positioning accuracy. The traditional calibration method is to expand the installation deflection angle error and lever arm error into state variables, the velocity difference under nonholonomic constraint (NHC) is taken as the measurement to establish the Kalman filter model for online calibration, but this method seriously depends on the sensor accuracy. In order to calibrate the installation errors of low-precision SINS, this paper improves NHC, adds the difference between the centripetal acceleration measured by odometer and calculated by SINS as measurement on the basis of the original measurement, and introduces a low-pass filter to reduce the interference of vehicle vibration on the accelerometer.The experimental results show that the proposed algorithm has better convergence of the installation deflection angle, and the error between the estimation results of the forward lever arm and the actual measurement results is less than 0.1m; After compensating the installation errors of the improved NHC, the maximum positioning errors in the east and north directions are reduced by 63.19% and 43.51% respectively, the height error is also greatly reduced, and the positioning accuracy of the integrated navigation system is significantly improved. [ABSTRACT FROM AUTHOR]

The mechanical structure of MEMS inertial sensors is a spring-mass-damper system, which is susceptible to interference near the resonant frequency. This work compared the resonance interference effects of mechanical vibrations and high frequency sound waves on inertial sensors. Three MEMS gyroscopes ADXRS620 were interfered by vibrations and sound waves near the resonant frequency. The error outputs increased linearly with the increase of interference intensity. In amplitude sweep experiment, the max error could reach 126.00 °/s, which seriously affect the normal operation of the gyroscope. The waveforms of gyroscope outputs under acoustic and vibration interference were almost coincident, which showed the similarity of acoustic and vibration interference. However, the MEMS accelerometer ADXL103 has a very different performance under acoustic and vibration interference, not only the resonant frequency and max error change, but also the error output and interference intensity are nonlinearly related. In amplitude sweep experiment, the max error could reach 9.78 m/s2. In addition, to produce the same effect on MEMS inertial sensors, the power required of vibration interference was much less than that of acoustic interference. The effectiveness against vibration interference of the filtering algorithm based on orthogonal demodulation was verified by testing the self-developed gyroscope. To the best of our knowledge, it is the first time that use the IQ two-channel orthogonal demodulation filtering algorithm to eliminate vibration interference. The elimination effect was over 97%. [ABSTRACT FROM AUTHOR]

This tutorial explores the many different ways that electrical resistance can be used to sense changes in the environment. Resistance-based sensors span a wide spectrum of transfer characteristics from the linear to the non-linear and from small, sub-Ohm changes to changes in resistance that span many orders of magnitude. Electronic interface circuits suitable for measuring these resistances range from those that convert sensor resistance to an analog voltage to those that generate a square wave whose frequency, duty cycle, or phase indicates resistance. In this tutorial, voltage dividers, Wheatstone bridges, current sources, and current pumps are discussed to represent the broad choices for converting resistance to an analog voltage. Oscillators, resistance-to-phase converters, direct interface circuits, and analog to digital converter circuits that convert signals directly from resistance to digital signals are also discussed. This tutorial seeks to provide the novice circuit or sensor designer with a place to start in selecting the right interface circuit for processing signals from resistance-based sensors. It also offers some additional interface circuit alternatives to the more experienced engineer. [ABSTRACT FROM AUTHOR]

MEMS accelerometers suffer from different problems such as random and systematic errors and the presence of outliers in the measurements. Proper calibration is necessary to obtain accurate results. Various methods exist but very few are designed for robust calibration. This paper explores methods for calibrating MEMS accelerometers in the presence of outliers. Optimization techniques based on Levenberg–Marquardt algorithm and linearization are used to solve the nonlinear calibration equations. Outlier detection methods such as RANSAC (random sample consensus), the Mahalanobis distance, and the median absolute deviation (MAD) are integrated within the optimization algorithm. Outlier scores are calculated and used to eliminate outliers or assign weights. The experimental results show that outlier detection and removal methods allow to achieve substantial improvement in the calibration process compared with non-robust methods. The median absolute deviation is the most effective in detecting outliers. In the presence of outliers, the average error for MAD and model-based method is 0.0136 (a.u.). For non-robust methods, the error is 3.82 (a.u.). For the Mahalanobis distance, the error is 1.8778 (a.u.). The results show that the calibration error is the smallest for the model—based method. The Mahalanobis distance presents higher calibration errors and is less capable of detecting outliers; however, it is still better than non-robust methods. Finally Levenberg–Marquardt algorithm presents slightly better results than linearization. [ABSTRACT FROM AUTHOR]

In the dynamic techniques for calibration of triaxial accelerometers, the sensor is held in motion during data acquisition; the acquired time-varying acceleration is in general more informative than the constant acceleration acquired in static calibration approaches. Dynamic methods, however, typically require complex and expensive calibration setups based on high-precision moving benches, to guarantee accurate a-priori knowledge of the acceleration applied to the sensors under test. This article presents a dynamic calibration procedure working with a very simple and inexpensive setup: a tilted, freely-rotating bench, made of a simple wheel mounted on a static support. The accelerometer is fastened to the bench, which is then manually set into rotation; the acceleration signal is then acquired while the bench is freely rotating. No a-priori knowledge is required about the bench rotation: the proposed calibration procedure estimates both the calibration parameters and the bench motion. To keep the setup as simple as possible, an effort has been made to minimize also the necessity of prior knowledge of the bench geometry: only the distance of the sensor from the rotation axis needs to be known, which can be easily obtained through direct measurement on the bench. The proposed calibration has been tested both on synthetic data, to prove the absence of estimation biases and to evaluate the potential accuracy of the estimated parameters, and on real data from a MEMS triaxial accelerometer, to assess the practical usability and measure the actual precision of the procedure. [ABSTRACT FROM AUTHOR]

In this study, we attempt to design a new square MEMS lead-free piezoelectric accelerometer that can be used simultaneously in a wider frequency range than has been available in the past and in low-G sensing environments with high accuracy and low transverse sensitivity. Low transverse sensitivity is an important factor for high-accuracy accelerometers. The device is designed through the use of a structural formula, and ANSYS software is used to simulate the vibration mode and resonance frequency of the device, after which the MEMS process is used to complete the fabrication of the devices. The performance of the proposed device indicated that the resonance frequency was 1740 Hz, with a simulation error of only 3.3%; the Z-axis sensitivity was 1.96 mV/g, and the transverse sensitivity ratio was 0.6%. To the best of the author’s knowledge, the transverse sensitivity obtained in the present work is the lowest compared to the published data and specifications for commercial devices. It was successfully applied to a low-frequency robotic arm and a high-frequency turbo pump, where the operational status of the devices was monitored in-situ with high accuracy. [ABSTRACT FROM AUTHOR]

Airborne transmittable diseases such as COVID-19 spread from an infected to healthy person when they are in proximity to each other. Epidemiologists suggest that the risk of COVID-19 transmission increases when an infected person is within 6 feet from a healthy person and contact between them lasts longer than 15 minutes (also called Too Close For Too Long (TC4TL). In this paper, we systematically investigate Machine Learning (ML) methods to detect proximity by analyzing publicly available dataset gathered from smartphones’ built-in Bluetooth, accelerometer, and gyroscope sensors. We extract 20 statistical features from accelerometer and gyroscope sensors signals and 28 statistical features of Bluetooth signal, which are classified to determine whether subjects are closer than 6 feet as well as the subjects’ context. Using machine learning regression, we also estimate the range between the subjects. Among the 19 ML classification and regression methods that we explored, we found that ensemble (boosted and bagged trees) methods perform best with accelerometer and gyroscope data while regression trees ML algorithm performs best with the Bluetooth signal. We further explore sensor fusion methods and demonstrate that the combination of all three sensors achieves a higher accuracy of range estimation than when using each individual sensor. We show that proximity (< 6ft or not) can be classified with 72%-90% accuracy using the accelerometer, 78%-84% accuracy using gyroscope sensor, and with 76%-92% accuracy with the Bluetooth data. Our model outperforms the current state-of-the-art methods using neural networks and achieved a Normalized Decision Cost Function (nDCF) score of 0.34 with Bluetooth radio and 0.36 with sensor fusion. [ABSTRACT FROM AUTHOR]

Due to the need for providing low-cost, small-size, and high-reliability calibration solution around the clock independently, the Micro-Electro-Mechanical System (MEMS) inertial measurement unit (IMU) is a hot topic currently. In this paper, we propose a new calibration method based on MEMS IMU with the assistance of the FOG IMU. Different from the traditional MEMS IMU calibration, a Four-Position-and-Three-Rotation (FPTR) calibration sequence is designed in the conditions of lacking turntable. The simulation result shows that all 24 calibration parameters converge through the calibration rotation scheme. An experiment result of qualitative analysis is presented, by comparing the traditional method and proposed calibration method, which reveals the repeatability and stability of the proposed method. In practical applications, this FPTR calibration method can effectively reduce the calibration time and improve the experimental efficiency. In summary, the proposed method is suitable for the mass production calibration of MEMS IMU, and possesses the ability to calibrate full parameters of MEMS IMU at once without turntable or other devices. The research outcomes can provide a practical basis for MEMS IMU plus FOG IMU calibration. [ABSTRACT FROM AUTHOR]

To date, the methodology to track and identify vertical movement from large-scale unstructured data sets is lacking. Here, we design and develop such a framework to accurately and systematically identify the sparse human vertical displacement activity typically buried into the predominantly horizontal mobility. Our framework uses sensor data from a barometer, accelerometer, and Wi-Fi scanner coupled with an extraction step involving a combination of feature engineering and data segmentation. This methodology is subsequently integrated into a machine-learning-based classifier to automatically distinguish vertical displacement activity—with 98% overall accuracy and a 92% F1-score—from its horizontal counterpart. We illustrate the potential of this framework by applying it to an unstructured large-scale data set associated with over 16,000 participants going about their daily activity in the city-state of Singapore. With the vertical movements of this large group uncovered, we can analyze the specific features of this activity class using its statistical distribution. This new knowledge would have significant ramifications for the architectural design of vertical cities. [ABSTRACT FROM AUTHOR]

Piezoelectric accelerometers present higher sensitivities and better signal-to-noise ratios in comparison with piezoresistive and capacitive accelerometers. A micro-electromechanical system (MEMS) piezoelectric accelerometer was designed and manufactured with four L-shaped beams covered by Lead Zirconate Titanate (PZT) thin film with $1~\mu \text{m}$ in thickness which was deposited using sol-gel methods. The micro devices, whose size and the optimized PZT distribution were determined by theoretical and simulation analysis, were fabricated on the silicon substrate using PZT films working in the ${d}_{31}$ mode. Different from the conventional vertical crossbeam accelerometers, the designed structure devotes to the larger area piezoelectric film laid on L-shaped supporting beams with higher energy transferring efficiency for sensing. Then the chips of accelerometer were fabricated, and the proposed sensor was packaged on the printed circuit board (PCB). In addition, the microsphere was integrated in the center of the beams. Finally, the properties of the accelerometer were measured by vibration measurement system. The measuring results show that the initial sensitivity of the accelerometer is 28.14mV/g at 500Hz. The proposed piezoelectric accelerometer has high sensitivity and low resonance frequency, and the frequency response on each beam is consistent, providing a new structure for high-performance accelerometer. [ABSTRACT FROM AUTHOR]

Microelectromechanical system (MEMS)-based resonant accelerometers have great application potential in navigation guidance. In the present work, a systematic theoretical model of a MEMS resonant accelerometer was proposed. The coupling effect of microleverage mechanisms on the proof mass and the resonator was analyzed by the displacement coordination method for the first time. The resonant frequency of the proof mass, the natural frequency of the resonator, and the sensitivity of the accelerometer were united into one model for the structural cooperative design. The analytical model was approved by finite element simulations and verification experiments, and the relative error of key parameters was less than 13%. The analysis of the accelerometer architecture revealed that 1) joints between beams and the microleverage arm should be on the same level, 2) the geometric dimensions of the input beam and pivot beam of the microleverage should be the same, and 3) the microleverage arm width should be sufficient to ensure the rigidity. The performance of the accelerometer was also tested. The bias instability of $1 \mu \text{g}$ (Allan variance), the stability of $2.12 \mu \text{g}$ ($1\sigma $ , 1 Hz date rate, one-hour length), and the nonlinearity of 16.3 ppm with the full range of ±20 g were achieved at room temperature. [ABSTRACT FROM AUTHOR]

The tunable coupling structure is proved to be an effective way to improve the adjustability of sensitivity of mode-localized sensors. However, the non-linearity introduced by the electrical coupling structure brings difficulties to its practical application. In this paper, a mechanical sensitivity adjustment method is proposed with lower output non-linearity of amplitude ratio and larger measurement range of acceleration. The proposed accelerometer is tested under closed-loop circuit to obtain 37% sensitivity adjustment range with 0.42% output non-linearity. The experimental results demonstrate 1.45/g initial sensitivity over 40g linear measurement range. The accelerometer also demonstrates $13.65 \mu \text{g}/\surd $ Hz noise floor and $22 \mu \text{g}$ bias-instability. The linear dynamic range is calculated to be 140.2 dB, which provides a very promising solution for acceleration measurement in high-end industrial applications. [ABSTRACT FROM AUTHOR]

Recently, there has been a notable shift to increasingly more sedentary lifestyles, yet it remains unclear whether inexpensive accelerometers, such as those contained in Actigraph devices, can provide consistent measurements of low magnitude accelerations. This study sought to explore the ability of two Actigraph models to differentiate low frequency oscillations, in reference to higher-end accelerometers with the idle sleep mode disabled (ISM) (Part 1) and enabled (Part 2) in a controlled environment. Eight GT9X, fifteen wGT3X-BT and two higher-end accelerometers (Triaxial ICP) were mounted to a 6-degree of freedom robot, which introduced frequencies ranging from 0.5-2.0 Hz (Part 1) and 0.5-4.0 Hz (Part 2). To compare the models, the minimum, maximum and range of outputs were calculated for each of the frequencies. Part 1 revealed that the Actigraph monitors were able to detect low frequency oscillations; the captured output was similar across the different Actigraph models but was significantly greater than the higher-end devices. Part 2 demonstrated that amplitudes greater than the described 40 m g (0.392 m/s2) threshold were required for the Actigraph monitors to wake up. This study demonstrates that the GT9X and wGT3X-BT Actigraph accelerometers can detect low magnitude movements when the ISM is disabled. [ABSTRACT FROM AUTHOR]

In this article, we review optical MEMS accelerometers with a particular focus on sensing techniques and interrogation methods. Optical accelerometers find use in various application domains ranging from microgravity to inertial navigation to vibration sensing. The performance of an accelerometer is quantified in terms of its range, bandwidth, sensitivity, and resolution. The combination of sensing technique and interrogation method determines the optical accelerometer’s performance. This article presents a classification in terms of guided-wave and free-space based optical sensing techniques used in acceleration measurement and their review. In free-space based sensing techniques, light propagating in free-space interacts with the mechanical structure resulting in modification of light properties at the receiver. In guided-wave based sensing techniques, light interaction with the mechanical structure is confined to the waveguide. Also, the different interrogation methods used in optical MEMS accelerometers are reviewed. The interrogation methods are classified as based on intensity modulation or frequency modulation of the optical signal received from sensor. In intensity-modulation based interrogation, light intensity at output is the measurand and, the cost and complexity of this class of methods is lower. In frequency-modulation based interrogation, the frequency or phase of the optical signal at the output is the measurand. Further, a high-resolution optical MEMS accelerometer based on waveguide Bragg gratings is described. A combination of free-space based sensing and intensity-modulation based interrogation methods will be suitable for consumer-grade accelerometer applications. For high-resolution applications like tactical and navigation grades, a combination of guided-wave sensing and frequency-modulation based interrogation methods would be appropriate. [ABSTRACT FROM AUTHOR]

Wearable sensor based Human Activity Recognition (HAR) has been widely used these years. This paper proposed a novel deep learning model for HAR using inertial sensors. First, a wearable device platform was developed with 6 inertial sensor units to collect triaxial acceleration signals during human movements, and the dataset of Command Actions of Traffic Police (CATP) was acquired. Then, a deep learning model named Bidirectional-Gated Recurrent Unit-Inception (Bi-GRU-I) was designed to improve the accuracy and reduce the amount of parameters. It is consisting of 2 Bi-GRU layers, 3 Inception layers, 1 Global Average Pooling (GAP) layer and 1 softmax layer. Finally, the comparing experiments with other methods were taken on 3 datasets: the self-collected CATP dataset, widely used Wireless Sensor Data Mining (WISDM) and University of California, Irvine (UCI-HAR) dataset. And the proposed method shows better performance and robustness. Moreover, the sensor configuration optimization was analyzed, and it shows that this method can also apply to the task using less sensor units. [ABSTRACT FROM AUTHOR]

Pedestrian positioning with wearable devices is a significant application of attitude tracking. It tracks the attitude (with heading angle being the most important part) of the device in real time and provides positioning services for users based on the information of step length provided by Pedestrian Dead-Reckon (PDR), which is a cheap and efficient positioning method at present. However, amid a train of positioning methods, the joint estimate of tracking is given by a train of methods based on the direction of gravity and the earths magnetic field direction. Considering the measurement of gravity that the gravity accelerometer is exposed to heavy noise due to the complex movement of human body during walking with uniform swing arm posture and forward speed, this paper proposed a novel estimate method based on the Kalman filter with multi-state constraints and the usage of low-cost sensors, which fulfills the estimation with the sequential observation of magnetic field. Compared with other related work, this method proposed in this paper eliminates the dependence on gravity direction, avoiding the influence of heavy noise caused by additional linear acceleration in motion state, and reduces the influence of insufficient observation when using magnetic field observation alone. The performance of the proposed method is evaluated by real-world experimentation results. [ABSTRACT FROM AUTHOR]

Pedestrian dead reckoning is a well-known approach for indoor navigation. There, the smartphone’s inertial sensors readings are used to determine the user position by utilizing empirical or bio-mechanical approaches and by direct integration. In this paper, we propose PDRNet, a deep-learning pedestrian dead reckoning framework, for user positioning. It includes a smartphone location recognition classification network followed by a change of heading and distance regression network. Experimental results using a publicly available dataset show that the proposed approach outperforms traditional approaches and other deep learning based ones. [ABSTRACT FROM AUTHOR]

This paper presents a novel urban transportation mode detection (TMD) method based on inertial and barometric data. To the best of the authors knowledge, this is the first time that a 3-axis accelerometer, a 3-axis gyroscope, and a barometer are combined for TMD purposes. One contribution was to build and share an optimized TMD dataset in terms of duration (44 hours), number of participants (34), sensor placements (on-foot, waist-attached and in the trouser’s pocket) and considered transportation modes (7). To infer transportation mode information, we tackle two classification scenarios: still, walk, bike, tramway, and bus, on the one hand, and still, walk, bike, elevator, stairs and public transport on the other hand. One major finding was that TMD classification accuracy has been widely overestimated by adopting the familiar randomized cross validation approach. According to the latter, the accuracy of our non-optimized classifiers ranged from 90% to 99%. However, few of them were accurate when the test samples were drawn from data of five unknown subjects. By selecting features in a personalized manner and by studying the importance of the placement of sensors, we realized that the latter significantly impact the performance of TMD models. Thus, we present here two robust TMD classifiers that show respectively an average accuracy of 75.63% for the first scenario and 79.41% for the second scenario using a foot-mounted sensor [ABSTRACT FROM AUTHOR]

Indoor navigation based on Pedestrian Dead Reckoning (PDR) is a highly researched topic in recent years. PDR is a two-dimensional dead reckoning technique that employs inertial sensors, such as those available in smartphones, to provide the user’s position. If the vertical position component is also required, it can be obtained from the smartphones’ barometer. However, unlike accelerometers, approximately, only 10% of the smartphones are equipped with barometers. In this paper, two methods for height difference estimation during stairs motion are proposed. Those are based only on the smartphones’ accelerometers and do not require any prior calibration. The proposed approaches are evaluated on different staircases with different smartphone locations on the user. Results show that the proposed approaches yield an average percentage error in the range of 2.5% – 13% when calculating height difference of the examined staircases. Thus, the proposed approach can be implemented alongside traditional PDR approaches to enhance its capabilities to 3D PDR during stairs motion. [ABSTRACT FROM AUTHOR]