Search

Gesture recognition by using Inertial Measurement Unit (IMU) sensors plays an important role in various Internet of Things (IOT) applications, e.g., smart home, intelligent medical system and so on. Traditional technologies usually utilize machine learning algorithms to train different gestures during the offline phase, then recognize the gesture during the online phase. However, such technologies cannot recognize these gestures without prior training. Even for the same gesture, with different gesture amplitude may result in unsuccessful recognition. Also if we change the person to perform the same gesture, the algorithms fails. In order to overcome these drawbacks, we propose an approach, which will be able to track the human body motion in realtime and also recognize complicated gestures. It utilizes the accelerometer information and proposes comprehensive localization algorithms for each deployed sensor attached on the human body. Then, it takes the correlation and limitation among body parts into account to recognize the gesture. Our experiments results show that, the successful recognition rate of our algorithm is 100%. Furthermore, any part of the human body can be well tracked, the tracking accuracy can reach 0.06m.

Road surface irregularities not only has an impact on consistency of road, but also affects driver safety, automobile mechanical stability, and fuel consumption. To order to evaluate road roughness and spot potholes, many solutions have been suggested to remotely track the road surface quality. Some of these methods use an embedded mobile accelerometer to take a crowd sensing view to feel the surface of road. Although crowd sensing has various advantages such as ubiquity and less cost, it has some sensitivity to the problems generated by structures made by human, actions of the driver, and characteristics of the road surface which cannot be identified road anomalies. For this purpose, this paper suggests method of deep learning which helps us (a) to identify the different types of surface of road and (b) to differentiate potholes automatically from destabilizations in the crowd sensing-based application sense arising from speed bumps or driver behavior. This paper study and implement different deep learning frameworks in particular: convolutionary neural networks, LSTM networks, and computational structures for reservoirs. The tests were performed using knowledge from the real world, and the results revealed good precision in solving both of the problems.

Seismic responses of utility tunnel-soil system were studied via shaking table model tests with considerations of two kinds of double box utility tunnels: with and without joint connections. These two testing utility tunnel models were made of galvanized iron wire and micro-concrete, and the ground was simulated by the dry standard sand through layered tamping treatment. The utility tunnel-soil system was subjected to horizontal vibration in uniaxial direction perpendicular to the longitudinal direction of tunnel model. Via instrumentations of earth pressure gauges, accelerometers and strain gauges, the earth pressure response, acceleration response and bending moment response were measured. The testing results show that the joint connections in the utility tunnel along the longitudinal direction play an important role in determining the characteristic of earth pressure response and bending moment response, whereas the effect of joint connections on acceleration response is less significant. In addition, the partition wall exhibits the consistent acceleration response with the side-wall of double box utility tunnel model under seismic condition. Based on the testing results, it is suggested that the joint connection should be taken reasonably into consideration during design and construction for engineering practice.

Cow behavior recognition systems support the assessment of cows’ condition by providing their behavior information. Accelerometers are particularly suited for a non-invasive solution for the development of these monitoring systems. They are cheap and simple in setting up and providing high-performance recognition when using machine learning algorithms. The activity complexity of animals brings challenges in real context applications because different behaviors may have similar acceleration data. The reason is that they contain similar gestures, for example, feeding and standing. In our previous work, we proposed a cows’ behavior classifier based on leg-mounted acceleration data. The distinguishing of similar behaviors, such as feeding and standing, is limited by the data. This study presents a new efficient cow behavior recognition system based on combining leg-mounted and collar-mounted accelerometers. Significantly, the acceleration data from these two sensors were synchronized. Therefore, we can substantially expand the amount of information for classification purposes. Our approach identifies four cow behaviors: walking, feeding, lying, and standing. Random Forest algorithm with our extracted features (root mean square, standard deviation, and mean) and 16-second data window (a sample/second) offer excellent performance when identifying all concerning behaviors: feeding (0.914 accuracy, 0.884 sensitivity, 0.956 positive predictive value), lying (0.998, 0.996, 1), standing (0.88, 0.928, 0.842), and walking (0.998, 0.996, 0.998). These performances are better than other existing works, especially in our experiments with free-grazing cows.

The proliferation of smartphones equipped with various sensors created a large user base, which includes robots (smartphone controlled) as well. The high processor enabled smartphones are used as software robots to track hour-to-hour activities of users using on-board motion sensors. The magnetometer, accelerometer, and gyroscope are non-permission based motion sensors. The logs of these sensors create discriminative patterns to identify users’ movements. This leverages an adversary to cause privacy threats using these non-permission based sensors without user consent. Recent research works also highlighted the privacy threats due to these motion sensors in smartphones. Therefore, in this paper, we present a potential misuse of technology where smartphones can be hacked by PIN Inference attack endangering the physical and software robots’ controlled user activities. We perform experiments to collect motion sensors’ data and infer 4-digit PIN of a smartphone. Our strategy is to detect every key pressed one by one constituting a 4-digit PIN. Our inference results are promising and prove that an adversary can make use of sensors’ data to infiltrate user’s sensitive information from smartphones. Our proposed model infers 84% of the PINs correctly within 40 attempts when tested on 50 PINs. Given real Android devices with different users, we can identify the PIN of smartphones by training our machine learning based model. We also compare our approach with existing state-of-the-art approaches to show the efficacy of our attack model.

In the last two decades, fall detection (FD) systems have been developed as a popular assistive technology. Such systems automatically detect critical fall events and immediately alert medical professionals or caregivers. To support long-term FD services, various power-saving strategies have been implemented. Among them, a reduced sampling rate is a common approach for an energy-efficient system in the real-world. However, the performance of FD systems is diminished owing to low-resolution (LR) accelerometer signals. To improve the detection accuracy with LR accelerometer signals, several technical challenges must be considered, including misalignment, mismatch of effective features, and the degradation effects. In this work, a deep-learning-based accelerometer signal enhancement (ASE) model is proposed to improve the detection performance of LR-FD systems. This proposed model reconstructs high-resolution (HR) signals from the LR signals by learning the relationship between the LR and HR signals. The results show that the FD system using support vector machine and the proposed ASE model at an extremely low sampling rate (sampling rate < 2 Hz) achieved 97.34% and 90.52% accuracies in the SisFall and FallAllD datasets, respectively, while those without ASE models only achieved 95.92% and 87.47% accuracies in the SisFall and FallAllD datasets, respectively. This study demonstrates that the ASE model helps the FD systems tackle the technical challenges of LR signals and achieve better detection performance.

Drowsiness detection is a crucial step for safe driving. A plethora of efforts has been invested on using pervasive sensor data (e.,g., video, physiology) empowered by machine learning to build an automatic drowsiness detection system. Nevertheless, most of the existing methods are based on complicated wearables (e.,g., electroencephalogram) or computer vision algorithms (e.,g., eye state analysis), which makes the relevant systems hardly applicable in the wild. Furthermore, data based on these methods are insufficient in nature due to limited simulation experiments. In this light, we propose a novel and easily implemented method based on full non-invasive multimodal machine learning analysis for the driver drowsiness detection task. The drowsiness level was estimated by self-reported questionnaire in pre-designed protocols. First, we consider involving environmental data (e.,g., temperature, humidity, illuminance, and further more), which can be regarded as complementary information for the human activity data recorded via accelerometers or actigraphs. Second, we demonstrate that the models trained by daily life data can still be efficient to make predictions for the subject performing in a simulator, which may benefit the future data collection methods. Finally, we make a comprehensive study on investigating different machine learning methods including classic `shallow' models and recent deep models. Experimental results show that, our proposed methods can reach 64.6% unweighted average recall for drowsiness detection in a subject-independent scenario.

Tracking ground support equipment (GSE) in a high accuracy manner is crucial for both airport safety and optimal management of airport assets but the related researches and products are scarce. Tracking a multi-carriage logistics train is obviously most challenging compared with other single-carriage GSE. In this paper, we design a real-time on-board positioning and heading system (ROPHS) to obtain the real-time status of a multi-carriage logistics train which consists of a powered leading vehicle and one or several non-powered trailing vehicles. The status includes: (a) the accurate positions and velocities of any points on this train, and (b) the alterable number and linking sequence of trailing vehicles at any time of a trip. Technically, the hardware of the system relies on real-time kinematic (RTK) to obtain geolocation of the leading vehicle, and on gyroscopes, magnetometers and accelerometers to obtain headings of all vehicles. A geometry based recurrence algorithm is afterwards presented to calculate the positions of any trailing vehicles. In the end, the multiple model based tracking algorithm is proposed to compute the precision-improved locations and real-time velocities of the whole train. Different from existing traditional GPS or RFID based positioning techniques, the proposed system can reach centimeter-level accuracy, which enables collisions detection, especially those latent collisions that cannot be easily monitored or foreseen by crews' visual inspection.

Driving is an integral component of many operational systems, and any small improvement in driving quality can have a significant effect on accidents, traffic, pollution, and the economy in general. However, making improvements is challenging given the complexity and multidimensionality of driving as a task. In this paper, we investigate the effectiveness of nudging to improve driving performance. In particular, we leverage a smartphone application launched by our industry partners to send three types of nudges through notifications to drivers, indicating how they performed on the current trip with respect to their personal best, personal average, and latest driving performance. We measure the resulting driving performance using telematics technology (i.e., real-time sensor data from an accelerometer, Global Positioning System (GPS), and gyroscope in a mobile device). Compared with the “no-nudge” control group, we find that personal best and personal average nudges improve driving performance by approximately 18% standard deviations of the performance scores calculated by the application. In addition, these nudges improve interaccident times (by nearly 1.8 years) and driving performance consistency, as measured by the standard deviation of the performance score. Noting that driving abilities and feedback seeking may vary across individuals, we adopt a generalized random forest approach, which shows that high-performing drivers who are not frequent feedback seekers benefit the most from personal best nudges, whereas low-performing drivers who are also frequent feedback seekers benefit the most from the personal average nudges. Finally, we investigate the potential mechanism behind the results by conducting an online experiment in a nondriving context. The experiment shows that the performance improvements are directly driven by the changes in participants’ effort in response to different nudges and that our key findings are robust in alternative (nondriving) settings. Our analysis further shows that nudges are effective when the variability in reference points is low, which explains why the personal best and personal average nudges are effective, whereas the last score nudge is not. This paper was accepted by Vishal Gaur, operations management.

Wearable activity recognition can collate the type, intensity, and duration of each child's physical activity profile, which is important for exploring underlying adolescent health mechanisms. Traditional machine-learning-based approaches require large labeled data sets; however, child activity data sets are typically small and insufficient. Thus, we proposed a transfer learning approach that adapts adult-domain data to train a high-fidelity, subject-independent model for child activity recognition. Twenty children and twenty adults wore an accelerometer wristband while performing walking, running, sitting, and rope skipping activities. Activity classification accuracy was determined via the traditional machine learning approach without transfer learning and with the proposed subject-independent transfer learning approach. Results showed that transfer learning increased classification accuracy to 91.4% as compared to 80.6% without transfer learning. These results suggest that subject-independent transfer learning can improve accuracy and potentially reduce the size of the required child data sets to enable physical activity monitoring systems to be adopted more widely, quickly, and economically for children and provide deeper insights into injury prevention and health promotion strategies.

Identifying vehicle maneuvers in the context of Connected Vehicles (CV) system brings huge potentials to enhance traffic safety. However, this process requires various advanced sensors, which are either available for luxury vehicles or expensive to install. Differently, smartphone is a more feasible choice with high penetration rate and various built-in sensors. Among the existing studies of applying the smartphone to detect vehicle maneuvers, most treated the detection as a classification problem without considering the real-world application. For example, the smartphone was fixed. Too many descriptive features were generated from the sensor data. To alleviate these problems, this paper developed a vehicle maneuvers detection system using a common smartphone with GPS, gyroscope, accelerometer, and magnetometer sensors. We first released the constraints on the smartphone's position through a coordination system reorientation method. Then, simply filtered sensor data were directly used. A stacked-LSTM model was built to detect the vehicle maneuvers considering the time-dependency of the sensor data. This paper compared the performance of the proposed system with previous studies and various machine learning methods, including LightGBM, K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Random Forest. Extensive experimental results indicated that the proposed system accurately detected different vehicle maneuvers with an average F1-score of 0.98, precision of 0.97, and recall of 0.98, which outperformed the counterparts. Moreover, the model can be easily transferred to different drivers and locations. The system is robust and suitable for the real-time application as it requires simple processing of smartphone sensor data.

INTRODUCTION: Recent advances in wearable technology allow for the development of wirelessly connected sensors to continuously measure vital parameters in the general ward or even at home. The present study assesses the accuracy of a wearable patch (Healthdot) for continuous monitoring of heartrate (HR) and respiration rate (RR).MATERIALS AND METHODS: The Healthdot measures HR and RR by means of chest accelerometry. The study population consisted of patients following major abdominal oncological surgery. The analysis focused on the agreement between HR and RR measured by the Healthdot and the gold standard patient monitor in the intensive and post-anesthesia care unit.RESULTS: For HR, a total of 112 h of measurements was collected in 26 patients. For RR, a total of 102 h of measurements was collected in 21 patients. On second to second analysis, 97% of the HR and 87% of the RR measurements were within 5 bpm and 3 rpm of the reference monitor. Assessment of 5-min averaged data resulted in 96% of the HR and 95% of the RR measurements within 5 bpm and 3 rpm of the reference monitor. A Clarke error grid analysis showed that 100% of the HR and 99.4% of the 5-min averaged data was clinically acceptable.CONCLUSION: The Healthdot accurately measured HR and RR in a cohort of patients recovering from major abdominal surgery, provided that good quality data was obtained. These results push the Healthdot forward as a clinically acceptable tool in low acuity settings for unobtrusive, automatic, wireless and continuous monitoring.

A recent survey of Europe’s highway infrastructure has concluded that almost half of Europe’s bridges are nearing the end of their design live. Work in the wider Structural Health Monitoring sector is aiming to develop reliable and cost-effective methods for verifying condition, remaining service life and safety of ageing structures. Most bridge condition assessment methods are based on deflection, acceleration or strain measurements. This paper looks at the possibility of using rotation measurements as a main parameter to identify damage. This study looks at numerical analyses of a moving point load on a one-dimensional bridge model to provide the theoretical basis of the proposed damage detection method. It is shown that when local damage occurs, even when it is remote from a sensor location, it results in an increase in the magnitude of rotation measurements. This study looks at how best to exploit this fact for damage detection. A number of damage scenarios, sensor locations, and load arrangements are investigated in this study and their influence on the ability of the algorithm to detect damage are reported.

This study presents a method for predicting location classes of a room such as a kitchen, and restroom, where a user is located by discovering location-specific sensor data motifs in sensor data observed by user’s sensor devices, such as smartwatch, without requiring labeled training data collected in a target environment. For example, we can observe similar waveforms corresponding to kitchen knife chopping actions using body-worn accelerometers in kitchens and can also observe similar sound features by active sound probing in bathrooms because of their water-resistant walls. This indicates that such location-specific sensor data motifs can be inherent information for location class prediction in almost every environment. This study proposes a novel method that automatically detects location-specific motifs from time series sensor data by calculating a score that represents the “location specificity” of each motif in a time series. Previous studies on location class prediction assume that location-specific sensor data are always observed in a room or use handcrafted rules and templates to detect location-specific sensor data resulting in difficulties in applying them to several realistic environments. In contrast, our method, named IndoLabel, can automatically discover short sensor data motifs, specific to a location class, and can automatically build an environment-independent location classifier without requiring handcrafted rules and templates. The proposed method was evaluated in real house environments using leave-one-environment-out cross-validation and achieved a state-of-the-art performance although labeled training data in the target environment was unavailable.

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.

The paper presents a research of angular orientation based on a Microelectromechanical System (MEMS) accelerometer by using machine learning (ML) and deep learning (DL) model with architectures of deep neural networks (DNN). In the industrial environment, Artificial Intelligence (AI) plays a crucial role in automation which is a potential solution for better performance of inclinometer. This research was carried out to apply this intelligent model on the Inertial Measurement Unit (IMU) to accomplish the angular position. The experiment shows that the ML model correctly learns the relationship between acceleration and tracking angles via polynomial regression with an R-square of 0.99. The employed DL model with 4 hidden layers of 10 neurons achieves an accuracy of 99.99 \% and almost non-error performance. The acceleration acquisitions were obtained from MEMS accelerometer LSM9DS1 at a frequency of 50 Hz via microcontroller STM32F401RE. The ML and DNN model were designed based on platform Tensorflow with high processing accuracy. The Pan-Tilt Unit was used as the angle reference for static and dynamic tests. The traditional technique is used for comparison as well as verification of the proposed models. DL model has better precision over the ML model due to its high structure level with updating weight and error optimization from the neural network structure. Meanwhile, MC shows more stable results in dynamic circumstances.

Balance ability is important for preschoolers' motor and physical development. Portable accelerometers can provide resolution tests and identification of preschoolers with balance defects.Despite previous studies on the balance measures of accelerometer tests, there is a lack of complete analyses for preschoolers aged 4-5 years. In this study, we aim to verify the reliability of measuring the static balance of preschoolers in this age range based on inertia sensors for the acceleration and angular velocity moduli.Thirty children wore an inertial sensor in the 5th lumbar vertebra and completed four tests, i.e., standing on a firm surface and on a foam surface with open and closed eyes. The standard deviation of the acceleration modulus and root mean square of the angular velocity modulus were calculated. The analysis was based on the intraclass correlation coefficient (ICC) to determine the internal consistency and feasibility.The ICC of the acceleration modulus was between 0.597 and 0.683 (P 0.01), and the test-retest reliability was medium. The ICC of the angular velocity modulus was between 0.683 and 0.812 (P 0.01, P 0.001), and the test-retest reliability was medium to good. The standard error of measurement (SEM) of the acceleration modulus was between 0.001591 and 0.007248 (g), and the SEM% was between 21.24% and 34.12%. The angular velocity modulus SEM values ranged from 1.296 to 3.441 (deg/s), and the SEM% ranged from 25.17% to 33.26%. The difference between the two tests was evenly distributed on both sides of the mean value, and the difference between the test results was within the consistency limit.Inertial sensors can be used to evaluate the static balance ability of preschoolers aged 4-5 years. Further, the angular velocity modulus is more reliable than the acceleration modulus.

The rapid development of Information-Centric Wireless Sensor Network (ICWSN) has solved the challenges of information transmission and processing caused by the accelerated growth of wearable devices and the wide deployment of the Internet of Things (IoT) recently. The privacy security is also a growing problem. The existing works use earphones, covert and user-friendly wearable devices, for user authentication. However, some of the earphone-based authentication solutions need to customize special earphones, which are not universal. Other solutions use microphones and speakers of earphones for authentication, which are susceptible to changes in the auricle’s internal environment, resulting in a decline in performance. To solve this problem, a new authentication solution based on the existing commercial earphones is proposed to authenticate a user by tapping on the earphone rhythmically. This rhythmic tap behaviour causes a change of the signal waveform of the built-in accelerometer in the earphone. Based on this, we design a pipeline to authenticate the user’s identity. We first design an event detection algorithm to segment the tap signal accurately. Then we use the global features calculated based on event detection algorithm and local features extracted from the Convolutional Neural Network (CNN) for building an authentication model using the Naive Bayes (NB) classifier. Finally, 20 users are recruited to evaluate the experiment, and the recognition accuracy reaches 98%. Moreover, we extend the experiment to prove that it has a good performance against the different attacks and is robust in different scenarios.

Background:This research aimed to determine whether patterns of temporal changes in activity levels can indicate postoperative complications following hepatectomy., Methods:Between December 2016 and December 2019, 147 patients wore an accelerometer to measure their physical activity levels after hepatectomy until postoperative day 7. Patterns of changes in activity levels were categorized as follows: upward slope type (n = 88), wherein activity levels gradually increased; bell curve type (n = 13), wherein activity levels initially increased but subsequently decreased; and flat type (n = 46), wherein there was no apparent increase in activity levels. Patient characteristics and postoperative complications were compared for each group., Results:Postoperative complications occurred in 4.5% of patients in the upward slope group, in 76.9% in the bell curve group, and in 65.2% in the flat group (p < 0.001). Surgical site infections (SSI), refractory pleural effusion, and ascites were more common in the bell curve group, while pneumonia was only observed in the flat group., Conclusion:SSI, pleural effusion, and ascites should be considered when previously increasing activity levels decline during the postoperative period. In addition, there is a high risk of SSI and pneumonia when activity levels do not increase at all after surgery.

Background: Commercially available wrist-mounted exercise monitors may offer objective data on disease and recovery. This study is the first to evaluate the potential of such devices in the assessment of frozen shoulder and the effects of treatment. Methods: Twenty-one patients with isolated, unilateral frozen shoulder wore a wrist-mounted accelerometer (Fitbit Fire II, Fitbit Inc. 2007, California, USA) on each wrist for two separate seven-day periods, one week before and six months after treatment. The monitors produced an activity count for each 24-hour period, accounting for all movements of the upper limb. Three values were calculated for each time period: (1) the mean activity count for each limb, (2) the total activity count for both limbs, and (3) an activity count ratio calculated by dividing the activity of the frozen limb by the unaffected limb. Constant score, American Shoulder and Elbow Surgeons, visual analog scale–pain, and range of movement were recorded before and after treatment. Results: Mean activity counts were significantly lower in the frozen shoulder limb than those in the unaffected limb over the initial seven-day period (6066 vs. 7516; P = .04). The activity count ratio significantly improved after treatment (0.83 vs. 096; p 0.01), whereas the mean total activity count remained similar before and after treatment (14915 vs. 12371; P = .18), demonstrating that activity transferred from the unaffected limb back to the previously frozen limb. Range of movement (P

Background Persons with Parkinson’s disease (PD) participate in low levels of physical activity. This has prompted interest in developing interventions targeting physical activity behavior in PD. However, the current cut-points to quantify moderate-to-vigorous physical activity (MVPA) developed for PD have been derived from a single, vertical axis using hip-worn accelerometers, and this cut-point may not be applicable for wrist-worn devices. Wrist-worn devices might improve accessibility and compliance with physical activity monitoring in PD. Research Question What is the relationship between wrist-based activity counts and energy expenditure during treadmill walking in persons with PD? Do cut-points for quantifying time spent in MVPA differ between persons with PD and controls matched by age and sex? Methods The sample included 26 persons with mild-to-moderate PD (Hoehn and Yahr stages 2–3) and 27 age- and sex-matched controls. Participants completed three, 6-minute bouts of walking on a treadmill at three increasing speeds. Vector magnitude was measured using ActiGraph GT3X+ accelerometer worn on the more affected side for persons with PD and the non-dominant side for controls. The rate of oxygen consumption, or energy expenditure, was measured using a portable, open-circuit spirometry system. Results Our results indicated a strong association between activity counts and energy expenditure for persons with PD and controls with R2 values of 0.94(0.07) and 0.95(0.06), respectively. Persons with PD had a cut-point of 2883(871) counts·min−1; this was significantly lower than the cut-point of 4389(1844) counts·min−1 for controls. Conclusion We generated a PD-specific cut-point for wrist-worn ActiGraph accelerometers among persons with PD, and this was lower than controls. This disease-specific cut-point may provide more accurate measurements of time spent in MVPA in PD.

We propose and develop a high-sensitivity fiber Bragg grating (FBG) accelerometer based on a flexible hinge for vibration sensing. The proposed method employs a 1 mm FBG that is inscribed pointwise using a femtosecond laser such that the resultant stress concentration at the FBG is improved significantly by reducing the distance between the two bonding points of the FBG structure. Through theoretical calculations and finite element simulations, the relationship between sensitivity and distance was found to be inversely proportional. Using this structure, the sensitivity was enhanced to 804 pm/g at 15 Hz and 30 Hz. The resonant frequency of the proposed accelerometer is 102 Hz, and the working frequency bandwidth is from 10 to 50 Hz. The performance of the accelerometer allows detection of vibrations in seismic explorations with high reliability.

The precise calculation of GRACE and GRACE-FO satellite orbits is reliant on knowledge of accurate non-gravitational accelerations acting on the spacecraft. These are measured by the on-board accelerometers that require a thermal environment stabilised to ∼ ± 0.1 °C per revolution. However, during periods of the GRACE mission with reduced thermal control, internal temperature variations reached up to 10 °C within a revolution, causing low-frequency and non-linear drifts in the accelerometer observations. Additionally, accelerometer bias drifts occurred throughout the GRACE mission as changes in the orientation of the orbital plane with respect to the Earth-to-Sun vector caused the satellites to absorb more or less solar energy. These temperature-induced drifts degrade the quality of mass change estimates, particularly during the latter half of the GRACE mission after thermal control of the satellites was terminated. We filter (in the frequency domain) the accelerometer observations to remove these low-frequency components ( f 0.045 mHz). The bias drift removed from the cross-track is then scaled to derive a thermally-based correction for the highly sensitive along-track observations. We then estimate temporal gravity fields using the ANU GRACE software, our filtered accelerometer observations and the range acceleration as the inter-satellite observation. The use of our thermally-corrected accelerometer measurements significantly improves the accuracy of both orbit modelling and gravity field estimation.

Use of a detailed parallel theoretical model has allowed the design of a high figure of merit and low cross sensitivity accelerometer based on fiber Bragg grating (FBG), and the good response characteristics have been demonstrated experimentally. The relationship between the sensitivity and resonant frequency is downplayed, the mechanism of expanding the frequency band and maintaining higher sensitivity has been analyzed. A FBG accelerometer with a good performance of higher main-axis sensitivity and lower cross sensitivity has been presented. The experimental results obtained indicate that the FBG accelerometer has a broad frequency range from 1 to 60Hz, the corresponding sensitivity range from 754.3 to 763.2pm/G, while the fluctuation is less than 3dB. The resonant frequency is 103Hz, and the corresponding cross sensitivity is less than -24.76dB (

The great popularity of smartphones, together with the increasingly important aim of providing context-aware services, has spurred interest in developing indoor tracking systems. Accurate tracking and localization systems are seen as key services for most context-aware applications. Research projects making use of radio signals detected by radio interfaces and the data captured by sensors commonly integrated in most smartphones have already shown promising and better results than location solutions based on a single data source. In this paper, we present a multi-sensor tracking system built by incrementally integrating state-of-the-art models of the Wi-Fi interface and the accelerometer, gyroscope and magnetometer sensors of a smartphone. Our proposal consists of a simple calibration phase of the tracking system, which involves enabling simultaneous data gathering from all three sensors and the Wi-Fi interface. Taking the Wi-Fi signal model as baseline, four different configurations are evaluated by incrementally adding and integrating the models of the other three sensors. The experimental results reveal a mean error accuracy of 60 cm in the case when the tracking system makes use of all four data sources. Our results also include a spatial characterization of the accuracy and processing power requirements of the proposed solution. Our main findings demonstrate the feasibility of developing accurate localization indoor tracking systems using current smartphones without the need for additional hardware.

This article proposes the mobile sensor-cloud (MSC) architecture and theoretically characterizes it. The existing literature on sensor-cloud (SC) assumes stationarity of nodes, which limits the use of the associated theoretical models in mobile platforms. This article attempts to address this research lacuna by proposing a model for enabling mobile sensor-as-a-service and analyzing its performance. An MSC, typically, orchestrates interoperation of multiple mobile sensors with stationary ones. Present day mobile devices are equipped with an assortment of built-in sensors such as accelerometer, gyroscope, GPS, and vibration. Considering the inherent mobility of these sensors by virtue of the mobility of their host platforms, it is important to characterize the MSC. We propose the concept of virtualization of mobile sensor nodes in the SC, which is one of the first attempts of its type. Additionally, in this article, we discuss the problems encountered due to the random mobility of mobile devices registered in the SC. Thereafter, we provide a possible solution to the problem by the reallocation of suitable physical mobile sensor nodes to an existing virtual sensor with available resources. Analytical results reveal that the lifespan of the network increases by 80%–90% in MSC, compared to the traditional SC. On the other hand, unlike traditional SC, twice the number of sensor-owners get the opportunity to participate in an MSC. However, in an MSC, an end-user pays similar amounts for an application involving SC. The mobility of devices creates opportunities for the sensor owners to appropriately utilize the available resources in the virtual sensors, which justifies the observed results.

TBM vibration is inevitable during excavation in hard rock or mixed face ground conditions (MFC) and is detrimental to equipment safety and environmental protection. However, the cutting-induced vibration can help clarify the TBM-ground interaction and provide a valuable ground identification approach. In the present investigation, a field measurement of TBM vibration in changing ground conditions was carried out with accelerometers mounted on the TBM bulkhead. The vibration characteristics and patterns under different ground conditions were compared by signal processing, and their relationships to the operating parameters were investigated. The results showed that the TBM dynamic response was highly dependent on geological conditions. In homogeneous soft ground (HSG), the magnitude of vibration was low and stable, while under the MFC, the high frequency and strong vibration occurred. The vibration waveform in the MFC had an apparent periodicity and was consistent with the cutter head rotation speed. In addition, the signal consisted of a series of periodical impulses with intervals, revealing the rock-cutting process. Since the TBM vibration was sensitive to ground changes, it could provide valuable and precise information about the cutting face ground conditions. This investigation of the relationship between the ground and vibration can provide a foundation for vibration-based ground identification, which is a potential application in simultaneously tracking geological conditions during tunneling.

A simulation study using an active vibration control system (AVCS) to reduce airframe vibrations of an unmanned compound helicopter is conducted. The present unmanned compound helicopter uses wings, propellers, and a lift-offset coaxial rotor system, with two blades per rotor for high-speed flight. The airframe vibration responses are reduced using the AVCS at 230 knots. The hub vibratory loads of a lift-offset coaxial rotor are calculated using a rotorcraft comprehensive analysis code, CAMRAD II. The obtained rotor vibratory loads excite the airframe structure, represented as a one-dimensional stick (elastic line) model. The finite element analysis software, MSC.NASTRAN, is used to model the structural dynamics and analyze the vibration response of the airframe. The AVCS in this work uses the Fx-LMS (Filtered-x Least Mean Square) algorithm to determine the vibration cancellation signal produced by the force generators, which has the same amplitude but opposite direction to the airframe vibration signal. Herein, the AVCS consists of four force generators and six accelerometers and the simulation framework is constructed using MATLAB Simulink. Ten AVCS simulations are conducted, considering the different locations and various directions of the vibration cancellation forces from the force generators. When AVCS is applied to the airframe at a flight speed of 230 knots, the 2/rev longitudinal and vertical vibration responses at the specified airframe positions, such as the remote cockpit device, wing root, and wing tip, are reduced by 81.39–99.93%, compared to the baseline results without AVCS application.

En este artículo se presenta la aplicación de métodos de análisis modal operacional (OMA, Operational Modal Analysis) con el objetivo de caracterizar los parámetros modales (v.g. frecuencias y modos de vibración) de una edificación. El estudio se realiza sobre un escenario real consistente en un edificio esencial. En concreto se emplea: el método de descomposición en el dominio de la frecuencia (FDD, Frequency Domain Decomposition) y su versión mejorada (EFDD, Enhanced-FDD). En una primera etapa, se lleva a cabo una evaluación estructural preliminar del edificio (empleando el método de inspección visual rápida, RVS), un levantamiento de dimensiones y ensayos en campo para la caracterización mecánica de sus componentes, todo ello con el propósito de conseguir un análisis modal convencional lo más fiable posible en términos de sus parámetros modales. Con base en este análisis modal, se diseña un plan de instrumentación con acelerómetros triaxiales de sistemas microelectromecánicos (MEMS); el proceso de instrumentación abarca tres etapas: la adquisición, el control y el almacenamiento de información. La principal contribución de este trabajo consiste en la evaluación de la aplicación de los métodos FDD y EFDD sobre un edificio esencial, con la particularidad del uso de vibraciones de microsismicidad para la identificación de parámetros modales. El análisis de los resultados obtenidos determina una frecuencia fundamental del edificio de 1.43 Hz, y evidencia un comportamiento modal no recomendado.

STUDY DESIGN Cross-sectional validation study. OBJECTIVES The performance of previously published physical activity (PA) intensity cutoff thresholds based on proprietary ActiGraph counts for manual wheelchair users (MWUs) with spinal cord injury (SCI) was initially evaluated using an out-of-sample dataset of 60 individuals with SCI. Two types of PA intensity classification models based on raw accelerometer signals were developed and evaluated. SETTING Research institutions in Pittsburgh PA, Birmingham AL, and Bronx NY. METHODS Data were collected from 60 MWUs with SCI who followed a structured activity protocol while wearing an ActiGraph activity monitor on their dominant wrist and portable metabolic cart which measured criterion PA intensity. Data was used to assess published models as well as develop and assess custom models using recall, specificity, precision, as well as normalized Mathew's correlation coefficient (nMCC). RESULTS All the models performed well for predicting sedentary vs non-sedentary activity, yielding an nMCC of 0.87-0.90. However, all models demonstrated inadequate performance for predicting moderate to vigorous PA (MVPA) with an nMCC of 0.76-0.82. CONCLUSIONS The mean absolute deviation (MAD) cutoff threshold yielded the best performance for predicting sedentary vs non-sedentary PA and may be used for tracking daily sedentary activity. None of the models displayed strong performance for MVPA vs non-MVPA. Future studies should investigate combining physiological measures with accelerometry to yield better prediction accuracies for MVPA.

A three-axis accelerometer sensor is incorporated in a leadless pacemaker (Micra™, MedtronicInc ) for rate adaptation. 3 sensor setpoints at lower rate (LR), activity of daily living rate (ADLR) and upper rate (UR) are used to convert detected acceleration to a desired rate response. We describe inappropriate rate acceleration at rest and during chest physiotherapy in a sedentary elderly patient with leadless pacemaker. The underlying mechanism and various programming options are discussed. This article is protected by copyright. All rights reserved.

One of the most important issues in civil engineering undergraduate education is response analysis of structures for earthquake-resistant design. Precautions that should be taken against seismic hazards not only include the design of earthquake-safe buildings but also monitoring the strength and dynamic characteristics of structures. Because this is an important topic in theoretical aspects of civil engineering education at the undergraduate level, it is essential to have opportunities to perform practices so students can better understand lectures. In the Gebze Technical University (GTU) Civil Engineering Department, we guide our undergraduate students to graduate as competent engineers who use practice tools. For this purpose, we designed an Arduino-based accelerometer device and its software toolkit, which records and visualizes structural vibration data in the process of learning the best practices of seismic safety. The device contains Arduino UNO board ADXL345 MEMS (microelectromechanical systems) Accelerometer and microSD card module. Both Arduino and Python open-source programming languages were implemented in the device. We have produced a total number of 15 accelerometers (named the ACCE_edu-Arduino based accelerometer for Civil Engineering education) integrating these cards and sensors that are widely used for vibration measurement and interpretation to record vibration data. Within the scope of “Python programming” lectures in the GTU undergraduate program, these toolkits will be used to obtain data that would be recorded, stored, visualized, and filtered using Python programming language, which provides a practical application in data processing.

Background Hip- and wrist-worn ActiGraph accelerometers are widely used in research on physical activity as they offer an objective assessment of movement intensity across the day. Herein we characterize and contrast key structured physical activities and common activities of daily living via accelerometry data collected at the hip and wrist from a sample of community-dwelling older adults. Methods Low-active, older adults with obesity (age 60+ years) were fit with an ActiGraph GT3X+ accelerometer on their nondominant wrist and hip before completing a series of tasks in a randomized order, including sitting/standing, sweeping, folding laundry, stair climbing, ambulation at different intensities, and cycling at different intensities. Participants returned a week later and completed the tasks once again. Vector magnitude counts/second were time-matched during each task and then summarized into counts/minute (CPM). Results Monitors at both wear locations similarly characterized standing, sitting, and ambulatory tasks. A key finding was that light home chores (sweeping, folding laundry) produced higher and more variable CPM values than fast walking via wrist ActiGraph. Regression analyses revealed wrist CPM values were poor predictors of hip CPM values, with devices aligning best during fast walking (R2 = 0.25) and stair climbing (R2 = 0.35). Conclusions As older adults spend a considerable portion of their day in nonexercise activities of daily living, researchers should be cautious in the use of simply acceleration thresholds for scoring wrist-worn accelerometer data. Methods for better classifying wrist-worn activity monitor data in older adults are needed.

Initialization is a fundamental task for the state estimation of multi-sensor integration. In this paper, we propose a novel approach to achieve a fast and accurate initialization for the integrated system. With the outputs from GNSS, the rough roll, pitch, yaw, and direction of gravity are calculated. Based on these initial states, the GNSS/INS fusion and Visual-Inertial Odometry (VIO) are simultaneously launched whenever the observations are available. Subsequently, a coarse scale is retrieved by assigning the initial states of GNSS/INS fusion to the VIO procedure. The scale can be utilized to get poses and points very close to the real world. Once the VIO estimation converges, the transformation parameters between the poses of VIO and GNSS/INS fusion are estimated via non-linear optimization. Thus, we can align these two trajectories and evaluate the scale error. Through six vehicular tests, we analyze the performance of the proposed method with different illumination, obstructions, and motion patterns, and compare the results of the proposed approach with the state-of-the-art algorithms. The results indicated that the Visual-Inertial parameters converge within 9 seconds in our method. The scale errors of both our method are mostly within 10%. The alignment errors are mostly at centimeter-level in our method, which is comparable to the existing approaches.

In this study we propose a portable tool to monitor and monitor patients at risk of Alzheimer's. The aim of this research was design and make a useful tool to determine the position and condition of Alzheimer's sufferers. The research stages consisted of designing hardware, software and testing the whole tool. The patient's condition find out by an accelerometer sensor was used which read the values of the x, y, and z axes to determine the condition of patient normal or falling. This tool also have a panic button which use by the patient if he did not know his position and forget the way to go home and SMS notification will be sent to the colleague when this button was pressed. The results of the test show the device can send an SMS notification of the patient's position when the supervisor sends a sms "GPS ON” and the navigation module can determine the location of the position. The distance between reading position of the GPS module to the test point in average was 2.39 meters. Generally speaking, the tool can working properly.

Smartphones became ubiquitous and are used by so many people, at least to know the driving directions to their destination. Smartphones come with rich embedded sensors (e.g., GPS, accelerometer, and camera) as well as built-in radios (e.g., Bluetooth, Wi-Fi, and Cellular), which both enable users to gather data and distribute it among people at any time or location. These features have come up with the mobile crowdsensing (MCS) development which can be used in a wide range of applications. In this paper, we introduce a complete mobile crowdsensing framework for road surface condition detection. Various modules have been addressed such as task management, data fusion, reputation scoring, incentive awarding, security and privacy, as well as discussing popular techniques and algorithms utilized in the proposed MCS framework modules. A prototype of the crowd sensing application is designed which is related to our framework. The proposed framework considers the data quality and trustiness between the users and the server as well.

Background: Observed associations of high-protein diets with changes in insulin resistance are inconclusive.Objectives: We aimed to assess associations of changes in both reported and estimated protein (PRep; PEst) and energy intake (EIRep; EIEst) with changes in HOMA-IR, glycated hemoglobin (HbA1c), and BMI (in kg/m2), in 1822 decreasing to 833 adults (week 156) with overweight and prediabetes, during the 3-y PREVIEW (PREVention of diabetes through lifestyle intervention and population studies In Europe and around the World) study on weight-loss maintenance. Eating behavior and measurement errors (MEs) of dietary intake were assessed. Thus, observational post hoc analyses were applied.Methods: Associations of changes in EIEst, EIRep, PEst, and PRep with changes in HOMA-IR, HbA1c, and BMI were determined by linear mixed-model analysis in 2 arms [high-protein-low-glycemic-index (GI) diet and moderate-protein-moderate-GI diet] of the PREVIEW study. EIEst was derived from energy requirement: total energy expenditure = basal metabolic rate × physical activity level; PEst from urinary nitrogen, and urea. MEs were calculated as [(EIEst - EIRep)/EIEst] × 100% and [(PRep - PEst)/PEst] × 100%. Eating behavior was determined using the Three Factor Eating Questionnaire, examining cognitive dietary restraint, disinhibition, and hunger.Results: Increases in PEst and PRep and decreases in EIEst and EIRep were associated with decreases in BMI, but not independently with decreases in HOMA-IR. Increases in PEst and PRep were associated with decreases in HbA1c. PRep and EIRep showed larger changes and stronger associations than PEst and EIEst. Mean ± SD MEs of EIRep and PRep were 38% ± 9% and 14% ± 4%, respectively; ME changes in EIRep and En% PRep were positively associated with changes in BMI and cognitive dietary restraint and inversely with disinhibition and hunger.Conclusions: During weight-loss maintenance in adults with prediabetes, increase in protein intake and decrease in energy intake were not associated with decrease in HOMA-IR beyond associations with decrease in BMI. Increases in PEst and PRep were associated with decrease in HbA1c.This trial was registered at clinicaltrials.gov as NCT01777893.

Rate‐adaptive single chamber pacemakers with accelerometer, closed loop stimulation (CLS), and remote monitoring functionality (Eluna 8 SR‐T, Biotronik, SE & Co, Germany) were implanted in 3 miniature donkeys with third‐degree atrioventricular block and syncope. After recovery, different pacemaker programming modes were tested at rest, during stress without physical exercise and during physical exercise. Pacing rates were compared to actual atrial rates and showed that CLS functionality allowed physiological heart rate adaptation. A transmitter installed in the stable provided wireless connection of the pacemaker to the internet. Home monitoring was activated which performed daily wireless transmission of pacemaker functional measurements to an online server allowing diagnosis of pathological arrhythmias and pacemaker malfunction from a distance. Closed loop stimulation and remote monitoring functionality resulted in nearly physiological rate adaptation and allowed remote “from‐the‐stable” patient follow‐up.

Voice assistant systems are convenient, but attackers can mimic users’ voices to access them and steal private information. We develop an authentication system that defends against acoustic attacks by using unique characteristics captured by the accelerometers in wearable devices.

In recent times, there has been an increase in the occurrence of disasters caused by a rapid increase in the aging of infrastructure. A structural health monitoring system can be utilized to detect severe damages in infrastructure and prevent accidents. In a previous study, the authors developed a method for damage detection in bridges using monitoring systems. However, the operational cost of the proposed method was high due to the requirement of multiple expensive accelerometers. The aim of this study is to reduce the number of accelerometers used in the aforementioned method by determining the optimal installation locations of the accelerometers. As a result of this study, optimal accelerometer installation locations for bridge damage detection, have been proposed.

Human motion is often tracked using three-dimensional video motion tracking systems, which have demonstrated high levels of validity. More recently, inertial measurement units (IMUs) have been used to measure human movement due to their ease of access and application. The purpose of this study was to systematically review the literature regarding the validity of inertial sensor systems when being used to track human motion. Four electronic databases were used for the search, and eleven studies were included in the final review. IMUs have a high level of agreement with motion capture systems in the frontal and sagittal planes, measured with root mean square error (RMSE), intraclass correlation coefficient, and Pearson's correlation. However, the transverse or rotational planes began to show large discrepancies in joint angles between systems. Furthermore, as the intensity of the task being measured increased, the RMSE values began to get much larger. Currently, the use of accelerometers and inertial sensor systems has limited application in the assessment of human motion, but if the precision and processing of IMU devices improves further, it could provide researchers an opportunity to collect data in less synthetic environments, as well as improve ease of access to biomechanically analyse human movement.

Background The usual accelerometry-based measures of physical activity (PA) are dependent on physical performance. We investigated the associations between PA relative to walking performance and the prevalence and incidence of early and advanced walking difficulties compared to generally used measures of PA. Methods Perceived walking difficulty was evaluated in 994 community-dwelling participants at baseline (age 75, 80, or 85 years) and 2 years later over 2 km (early difficulty) and 500 m (advanced difficulty). We used a thigh-mounted accelerometer to assess moderate-to-vigorous PA, daily mean acceleration, and relative PA as movement beyond the intensity of preferred walking speed in a 6-minute walking test (PArel). Self-reported PA was assessed using questionnaires. Results The prevalence and incidence were 36.2% and 18.9% for early and 22.4% and 14.9% for advanced walking difficulty, respectively. PArel was lower in participants with prevalent (mean 42 [SD 45] vs 69 [91] min/week, p < .001) but not incident early walking difficulty (53 [75] vs 72 [96] min/week, p = .15) compared to those without difficulty. The associations between absolute measures of PA and incident walking difficulty were attenuated when adjusted for preferred walking speed. Conclusions The variation in habitual PA may not explain the differences in the development of new walking difficulty. Differences in physical performance explain a meaningful part of the association of PA with incident walking difficulty. Scaling of accelerometry to preferred walking speed demonstrated independence on physical performance and warrants future study as a promising indicator of PA in observational studies among older adults.

To consider the influence of the quality of the accelerometer and the installation position on the test frequency, a test scheme was designed to identify the first-order bending frequency with free and cantilever beams as well as the first-order torsion frequency with free plates. According to the energy method and mode shape, the correction relationship between the elastic modulus and the first-order bending frequency of free and cantilever beams is derived alongside the relationship between the shear modulus and first-order torsion frequency of a free plate. These relationships are suitable for informing the influence of the accelerometer mass and installation. The first-order bending frequency and the first-order torsion frequency of Sitka spruce beams, slabs, and oriented strand board have been measured along the length, which verifies the effectiveness of the derived relation in testing the elastic modulus and shear modulus of wood materials. Taking the ratio m/M between the mass of the accelerometer and the mass of the test specimen as a parameter, the scope of application of the modified relation is discussed.

There is evidence that regular physical activity (PA) is associated with lower risk of non-communicable diseases, increased health status and the overall quality of life. Therefore, it is important to understand key factors that can influence PA levels and behaviours. There are several studies that investigated the determinants of PA in adults, however, they did not include environmental, behavioural, and sociodemographic factors to determine objectively measured PA behaviour. The aim of this study was to assess the association between objectively measured PA intensity of walking and sociodemographic, environmental and behavioural factors and investigate the most significant factors associated with the intensity of walking. The study included 125 randomly selected participants in Kaunas, Lithuania. Participants filled out a formalized questionnaire about sociodemographic, health-related, behavioural, and environmental factors. The intensity of PA of walking was measured using accelerometer and GPS (Global Positioning System) sensors. The results showed that the highest intensity of PA was found among the youngest age group, women, adults with normal weight and those with the lowest prevalence of chronic diseases including hypertension. In the multivariate regression models, age, noise level, and walking time remained significant covariates of the intensity of PA. Our study adds evidence on the separate and combined effects of sociodemographic, environmental, and behavioural factors on the intensity of PA during walking among adults in an urban environment using GPS and accelerometer sensors. Future studies should focus on additional factors influencing PA intensity that were not investigated in this study and explore other PAs.

This work presents the in-house design, characterization, and development of an integrated PZT-based piezoelectric MEMS vibration sensor module and its usage in surface roughness prediction of fused silica substrate in a CNC polishing machine. The differential electrical configuration of the MEMS device ensures enhanced acceleration sensitivity to out-of-plane motion, two orders larger than in-plane sensitivity. The Thin-film Piezo on Silicon (TPoS) accelerometer with a PZT layer was fabricated using photolithography, frontal etching, and backside release. A fully integrated module comprising MEMS, amplifiers, and a microcontroller was designed and characterized in a shaker system. The designed module provides a sensitivity of $8.12~mV/{g}$ and a resolution of $5.8~m{g}/\sqrt {Hz}$ . The feed rate, compared to the other polishing parameters, shows the highest correlation with surface roughness. The module’s acceleration-time data were processed by Linear Discriminant Analysis (LDA) and given as input to the neural network together with the feed rate of the polishing spindle. Back Propagation Neural Network (BPNN) algorithm was used to train the machine learning model. The designed module is cost-effective compared to the commercial sensor. The key performance indicators of roughness prediction for the module and the commercial sensor were calculated, showing a Mean Absolute Percentage Error (MAPE) of 6.45% and 5.32%, respectively.

This paper discussed the application of health monitoring systems to 20th-century historic buildings. Natural disasters are major threats to monuments. They are often seismically vulnerable and require interventions. However, taking into account their historic and cultural values, it is appropriate to observe long-term behaviour before making a decision on intervention schemes. To this aim, health monitoring is considered an effective approach. In recent years, MEMS (micro-electromechanical systems) accelerometers have been attracting attention for their convenience and efficacy. Nonetheless, the reliability of MEMS accelerometers still needs to be examined for the monitoring of monuments as sufficient research contributions have not been made. This paper presented two case studies that were monitored by means of MEMS accelerometers. They were masonry structures positioned in seismic-prone regions in Japan. A number of earthquakes were detected by the accelerometers during one year of monitoring. To examine the accuracy of the adopted MEMS accelerometers, dynamic identification tests were conducted using high-sensitivity strain-gauge accelerometers and servo velocity meters. Based on responses obtained from the tests, numerical simulation was performed. Nonlinear static analysis was performed. The numerical simulation permitted the comparison of reliability among sensors and test types. This paper provided suggestions for the dynamic identification tests of heritage structures.

Harbor caissons are regarded as one of the most critical infrastructures, and any disruption in their operation would have dramatic economic consequences. The need for a health monitoring system for these structures is thus evident. The primary aim of this study is to develop a structural health monitoring (SHM) system to effectively detect damage in harbor caissons. To this end, through performing dynamic analysis, acceleration signals are extracted from the locations in the model corresponding to the presumed accelerometer's placement in an actual structure. Three levels of white Gaussian noises are added to the original signal to simulate the ambient noise. Using the acquired signals, the damage-sensitive features in both frequency and time domains are determined in the structure. In this study, leveraging Principal Component Analysis (PCA) and Support Vector Machine (SVM), the damage-sensitive features associated with the damaged and undamaged structure are reduced in dimensionality and classified. Hence, the introduced system can assess the state of the structure based on the input signal from the accelerometer and detect possible damage and its severity. The effectiveness of the proposed system in detecting damage is shown using the numerical model of a real harbor caisson.