Search

In this paper, we present an occupant localization approach through sensing footstep-induced floor vibrations. Occupant location information is an important part of many smart building applications, such as energy and space management in a personal home or patient tracking in a hospital room. Adoption of current occupant location sensing approaches in smart buildings (e.g., camera, radio frequency (RF), mobile devices, etc.) is often limited due to the maintenance, installment, and calibration requirements of these sensing systems. To overcome these limitations, we introduce a new approach to use footstep-induced structural vibration for step-level occupant localization. The intuition behind this approach is that footsteps induce floor vibrations which are received in different vibration sensor locations at different times. This paper focuses on localizing a single occupant within each sensing range. To localize the footsteps, we utilize the time differences of arrival (TDoA) of the footstep-induced vibrations. However, this approach involves two main challenges: (1) the vibration wave propagation in the floor is of dispersive nature (i.e., different frequency components travel at different velocities) and (2) due to floor heterogeneity, these wave propagation velocities vary in different structures as well as in different locations in a structure. These issues lead to large localization inaccuracies or calibration requirements. To address dispersion challenge, we present a decomposition-based dispersion mitigation technique which extracts specific components (which have similar propagation characteristics) for localization. To address velocity variations, we introduce an adaptive multilateration approach that employs heuristics to constrain the search space and robustly locate the footsteps when the propagation velocity is unknown. Constraining the search space overcomes the additional complexity which is resulted from adding an unknown variable (propagation velocity). We evaluated our approach using the field experiments in 3 different types of buildings (both commercial and residential) with human participants. The results show an average localization error of 0.34 m, which corresponds to a 6X reduction in error compared to a baseline method. Furthermore, our approach resulted in standard deviation of as low as 0.18 m, which corresponds to a 8.7X improvement in precision compared to the baseline approach.

Many human activities induce excitations on ambient structures with various objects, causing the structures to vibrate. Accurate vibration excitation source detection and characterization enable human activity information inference, hence allowing human activity monitoring for various smart building applications. By utilizing structural vibrations, we can achieve sparse and non-intrusive sensing, unlike pressure- and vision-based methods. Many approaches have been presented on vibration-based source characterization, and they often either focus on one excitation type or have limited performance due to the dispersion and attenuation effects of the structures. In this paper, we present our method to characterize two main types of excitations induced by human activities (impulse and slip-pulse) on multiple structures. By understanding the physical properties of waves and their propagation, the system can achieve accurate excitation tracking on different structures without large-scale labeled training data. Specifically, our algorithm takes properties of surface waves generated by impulse and of body waves generated by slip-pulse into account to handle the dispersion and attenuation effects when different types of excitations happen on various structures. We then evaluate the algorithm through multiple scenarios. Our method achieves up to a six times improvement in impulse localization accuracy and a three times improvement in slip-pulse trajectory length estimation compared to existing methods that do not take wave properties into account.

We present a data fusion approach for enabling data-driven rail-infrastructure monitoring from multiple in-service trains. A number of researchers have proposed using vibration data collected from in-service trains as a low-cost method to monitor track geometry. The majority of this work has focused on developing novel features to extract information about the tracks from data produced by individual sensors on individual trains. We extend this work by presenting a technique to combine extracted features from multiple passes over the tracks from multiple sensors aboard multiple vehicles. There are a number of challenges in combining multiple data sources, like different relative position coordinates depending on the location of the sensor within the train. Furthermore, as the number of sensors increases, the likelihood that some will malfunction also increases. We use a two-step approach that first minimizes position offset errors through data alignment, then fuses the data with a novel adaptive Kalman filter that weights data according to its estimated reliability. We show the efficacy of this approach both through simulations and on a data-set collected from two instrumented trains operating over a one-year period. Combining data from numerous in-service trains allows for more continuous and more reliable data-driven monitoring than analyzing data from any one train alone; as the number of instrumented trains increases, the proposed fusion approach could facilitate track monitoring of entire rail-networks.

Collecting vibration data from revenue service trains could be a low-cost way to more frequently monitor railroad tracks, yet operational variability makes robust analysis a challenge. We propose a novel analysis technique for track monitoring that exploits the sparsity inherent in train-vibration data. This sparsity is based on the observation that large vertical train vibrations typically involve the excitation of the train’s fundamental mode due to track joints, switchgear, or other discrete hardware. Rather than try to model the entire rail profile, in this study we examine a sparse approach to solving an inverse problem where (1) the roughness is constrained to a discrete and limited set of “bumps”; and (2) the train system is idealized as a simple damped oscillator that models the train’s vibration in the fundamental mode. We use an expectation maximization (EM) approach to iteratively solve for the track profile and the train system properties, using orthogonal matching pursuit (OMP) to find the sparse approximation within each step. By enforcing sparsity, the inverse problem is well posed and the train’s position can be found relative to the sparse bumps, thus reducing the uncertainty in the GPS data. We validate the sparse approach on two sections of track monitored from an operational train over a 16 month period of time, one where track changes did not occur during this period and another where changes did occur. We show that this approach can not only detect when track changes occur, but also offers insight into the type of such changes.

We explore a data-driven approach for monitoring rail infrastructure from the dynamic response of a train in revenue-service. Presently, track inspection is performed either visually or with dedicated track geometry cars. In this study, we examine a more economical approach where track inspection is performed by analyzing vibration data collected from an operational passenger train. The high frequency with which passenger trains travel each section of track means that faults can be detected sooner than with dedicated inspection vehicles, and the large number of passes over each section of track makes a data-driven approach statistically feasible. We have deployed a test-system on a light-rail vehicle and have been collecting data for the past two years. The collected data underscores two of the main challenges that arise in train-based track monitoring: the speed of the train at a given location varies from pass to pass and the position of the train is not known precisely. In this study, we explore which feature representations of the data best characterize the state of the tracks despite these sources of uncertainty (i.e., in the spatial domain or frequency domain), and we examine how consistently change detection approaches can identify track changes from the data. We show the accuracy of these different representations, or features, and different change detection approaches on two types of track changes, track replacement and tamping (a maintenance procedure to improve track geometry), and two types of data, simulated data and operational data from our test-system. The sensing, signal processing, and data analysis we propose in the study could facilitate safer trains and more cost-efficient maintenance in the future. Moreover, the proposed approach is quite general and could be extended to other parts of the infrastructure, including bridges.

We introduce a physics-guided signal processing approach to extract a damage-sensitive and domain-invariant (DS & DI) feature from acceleration response data of a vehicle traveling over a bridge to assess bridge health. Motivated by indirect sensing methods' benefits, such as low-cost and low-maintenance, vehicle-vibration-based bridge health monitoring has been studied to efficiently monitor bridges in real-time. Yet applying this approach is challenging because 1) physics-based features extracted manually are generally not damage-sensitive, and 2) features from machine learning techniques are often not applicable to different bridges. Thus, we formulate a vehicle bridge interaction system model and find a physics-guided DS & DI feature, which can be extracted using the synchrosqueezed wavelet transform representing non-stationary signals as intrinsic-mode-type components. We validate the effectiveness of the proposed feature with simulated experiments. Compared to conventional time- and frequency-domain features, our feature provides the best damage quantification and localization results across different bridges in five of six experiments., To appear in Proc. ICASSP2020, May 04-08, 2020, Barcelona, Spain. IEEE

We have deployed our smart sensing system in a real commercial farm complex for at least 17 months. One of critical factors to success of the sensing system deployment is the resilience and fault tolerance of the system in a harsh environment with unreliable infrastructure and limited access. Power interruptions and intermittent connectivity is not uncommon. Sensors must work even submerging in animal excretion. Correct and continuous streams of sensor data is essential to our smart farming analytics. To make our sensing system sustain such challenges, we have designed and implemented the system with a capability of self-rejuvenation to ensure system liveness. We also equip it with our anomaly detection system to examine small sensor connectivity logs in order to identify potential faulty or deteriorating sensors or external event abnormality with minimal manual intervention.

In this paper, we present a physics-guided model transfer approach for robust human gait monitoring across various structures using footstep-induced structural vibrations. Understanding and characterizing the structural vibration caused by human footsteps provides insight about their gait, which is important in smart healthcare applications. Compared to current sensing approaches for gait monitoring (e.g., pressure-based, and mobile-based, and vision-based sensing), vibration sensing is sparse, non-intrusive, and causes less privacy concerns. Conventionally, finding gait-related information using structural vibrations occurs through data-driven inverse modeling with supervised learning, which extracts information about human gaits based on the vibration measurements and corresponding labels. However, the measurements are also affected by the structure properties; therefore, the models developed in one specific structure for mapping the measurements to labels are not applicable to other structures. This results in extensive calibration and training requirements. To overcome these challenges, we present a physicsguided model transfer approach that projects the vibration data into a space where the effect of structural properties is minimized. To this end, we first analytically show that the structural effect in the signal is correlated with the Maximum-Mean-Discrepancy (MMD) data distribution distance across various structures and then, minimize the MMD to find a projected space with reduced structural effect. This reduced structural effect enables us to train a footstep model in one structure (where labeled data is available) and use this model for labeling the samples in the other structures with no labeled data. In field experiments across two types of structures, our approach has achieved 0.95 and 0.97 F1-scores for footstep detection, which is up to a 9X improvement when compared to the baseline approaches.

In this paper we present a source separation approach which enables occupant gait monitoring for multiple concurrent walkers using footstep-induced floor vibration sensing. Occupant gait monitoring is a critical component of human-centric smart home applications like building security, gait health monitoring, and energy management. Current approaches such as vision sensing, wearables, and pressure sensors are limited in real-world deployments due to requirements of line-of-sight, the need for users to wear/carry a device at all times, and dense sensor deployment. Floor vibration sensing overcomes these limitations due to passive, sparse, and non-intrusive sensing, but prior works are limited to “single walker” scenarios. Real-world gait monitoring, however, typically involves persons walking in groups of two or more people (i.e., “multiple walker” scenarios). To this end, we introduce a frequency domain-based signal separation approach that enables gait monitoring in “multiple walker” scenarios by isolating the response from each concurrent walker. The insight is that the overall vibration response is the sum of each individual response signal at an unknown scale (amplitude) and time shift. We validate our approach with real-world walking experiments with two people walking side-by-side and correctly identify the order of the walkers with 88% accuracy (1.8X improvement over a baseline approach).

We propose a vehicle-vibration-based indirect structural health monitoring (SHM) framework that uses acceleration signals collected from within a moving vehicle to identify global modal and structural parameters of a full-scale and in-service bridge. Motivated by many benefits of indirect sensing methods, such as low-cost, low-maintenance and no interruption to traffic, researchers have in the past presented different algorithms and evaluated them on several simulation and lab-scale datasets. However, the uncertainties of the real-world vehicle-bridge interaction system and limited training data may cause previous methods to fail on full-scale bridges. To address these uncertainties, we 1) cast the vehicle-bridge interaction system as a linear time-varying Gaussian statespace model, which is not only able to estimate unobserved bridge responses but also able to add a stochastic process for modeling uncertainties, and 2) propose a hybrid algorithm that uses non-linear least squares and the expectation-maximization algorithm to estimate modal and structural parameters of the bridge using partially observed data (only the vehicle’s dynamic response is observed). We conducted field experiments on a steel truss bridge carrying two rail lines across the Monongahela River in Pittsburgh, Pennsylvania. For estimating the damage that is simulated by placing stationary trains on the bridge, our proposed approach has a 36.3% error reduction compared to a fully data-driven method. The results show that our proposed algorithm provides a potentially practical approach for continuous monitoring of in-service bridges.

Automated structural damage diagnosis is important for improving the efficiency of disaster responses and rehabilitation. In conventional data-driven frameworks using machine learning or statistical models, structural damage diagnosis models are often constructed in a manner of supervised learning. The supervised learning utilizes collected historical structural sensing data and respective damage states (i.e., labels) for each building to learn the building-specific mapping between them. However, in post-earthquake scenarios, historical data with labels are often not available for many buildings in the affected area, which makes it difficult to construct a damage diagnosis model. Directly using the historical data from other buildings to construct a damage diagnosis model for the target building would lead to inaccurate results. This is because data-driven models assume that the data distribution of training dataset (from other buildings) is consistent with that of test dataset (from the target building), while in practice each building has unique physical properties and thus unique data distribution. To this end, we introduce a new modeling framework to transfer the knowledge learned from other buildings to diagnose structural damage states in the target building. This framework is based on an adversarial domain adaptation approach that extracts domain-invariant feature representations of data from different buildings. The feature extraction function is trained in an adversarial way, which ensures the extracted feature distributions are robust to changes in structures while being predictive of the damage states. With the extracted domain-invariant feature representations, the data distributions become consistent across different buildings. We finally adopt the damage diagnosis model learned from other buildings datasets to infer the damage state of the target building. We evaluate our proposed framework using data collected from 5 different buildings subjected to 40 different earthquakes. We transfer the model learned from multiple buildings to diagnose a new building subjected to new earthquakes. The results show an up to 33.8% improvement in damage detection accuracy compared to conventional methods.

We present DR-Train, the first long-term open-access dataset recording dynamic responses from in-service light rail vehicles. Specifically, the dataset contains measurements from multiple sensor channels mounted on two in-service light rail vehicles that run on a 42.2-km light rail network in the city of Pittsburgh, Pennsylvania. This dataset provides dynamic responses of in-service trains via vibration data collected by accelerometers, which enables a low-cost way of monitoring rail tracks more frequently. Such an approach will result in more reliable and economical ways to monitor rail infrastructure. The dataset also includes corresponding GPS positions of the trains, environmental conditions (including temperature, wind, weather, and precipitation), and track maintenance logs. The data, which is stored in a MAT-file format, can be conveniently loaded for various potential uses, such as validating anomaly detection and data fusion as well as investigating environmental influences on train responses., Design Type(s)time series design • observation design • data collection and processing objectiveMeasurement Type(s)public infrastructure • positionTechnology Type(s)Accelerometer • GPS navigation systemFactor Type(s)temporal_interval • temperature • wind • weather • precipitationSample Characteristic(s)Pittsburgh • anthropogenic habitat Machine-accessible metadata file describing the reported data (ISA-Tab format)

This paper presents an approach that estimates occupants temporal gait parameters (e.g. step time, stride time, and initial loading response time) through structural floor vibrations while adapting to varying structural conditions. Continuous monitoring of these temporal gait parameters is a key component for elderly fall risk assessment and diagnosing injuries and gait abnormalities in home healthcare. A primary research challenge with using floor vibration sensing is that real-world deployments are often conducted in existing structures where little is known about the underlying structural properties. As such, prior works utilizing floor vibrations combine data-driven approaches with key heuristic observations to achieve accurate results, but often fail to adapt to varying structural conditions without re-training. To overcome this challenge, we present our approach, which extracts dynamic properties of the structure from footstep-induced vibrations and uses those properties to update gait parameter estimation models. By passively learning the dynamic properties of the structure, our approach enables gait parameter estimation that is robust to structural changes. We evaluate our approach by conducting real-world walking experiments on a test bed at Carnegie Mellon University.

Sensor-based occupant tracking has the potential to enhance knowledge of the utilization of buildings. Occupancy-tracking strategies using footstep-induced floor vibrations may be beneficial for thermal-load prediction, security enhancement, and care-giving without undermining privacy. Current floor-vibration-based occupant-tracking methodologies are based on data-driven techniques that do not include a physics-based model of the structural behavior of the floor slab. These techniques suffer from ambiguous interpretations when signals are affected by complex configurations of structural and non-structural elements such as beams and walls. Using a physics-based model for data-interpretation enables deployment of sparse number of sensors in contexts of non-uniform structural configurations. In this paper, an application of physics-based data interpretation using error-domain model falsification (EDMF) is presented to track an occupant within an office environment through footstep-induced floor vibrations. EDMF is a population-based approach that incorporates various sources of uncertainty, including bias, arising from measurements and modeling. EDMF involves the rejection of simulated model responses that contradict footstep-induced floor vibration measurements. Thus, EDMF provides a set of candidate locations from an initial population of possible occupant locations. A sequential analysis that accommodates information from previous footsteps is then used to enhance candidate locations and identify trajectories among candidates. In this way, incorporating structural behavior in interpreting vibration measurements induced by occupant footsteps has the potential to identify accurately the trajectory of an occupant in buildings with complex configurations, thereby providing tracking information without undermining privacy.

Mobile crowd sensing (MCS) enables many smart city applications (e.g., transportation monitoring/management, environmental monitoring, etc.). Recently, MCS systems built on non-dedicated vehicular platforms like taxis have become popular due to their large-scale coverage and low-cost deployment and maintenance. However, the goal of MCS may be inconsistent with the goal of vehicles. For example, MCS expects to get large and balanced sensing coverage over the city, while the taxis gather in busy areas to search for new ride requests. This inconsistency between the goals of MCS and vehicles results in a low sensing coverage and decreases the quality of the collected information. To address this inconsistency and optimize the sensing coverage, this paper presents an incentivizing system to optimize the sensing coverage of the sampled data. Key challenges to resolving this inconsistency include limited budget constraining the ability to incentivize more vehicles and complicate vehicle and trajectory selection problem making it difficult to obtain the incentivizing strategy. To address these challenges, we design a customized incentive by combining monetary incentives and potential ride request at the destination to reduce the cost of incentivizing vehicles and utilize the budget efficiently. Meanwhile, we formulate the problem of incentivizing trajectory planning as a non-linear multiple-choice knapsack problem, and propose a heuristic algorithm to approximate the optimal incentivizing strategy. The experiments based on the real-world data show that our system achieves up to 26.99% improvement in the sensing coverage compared to benchmark methods.

Track geometry is one of the most important health indices in the maintenance of rail tracks. Visual inspection and inspection using a track-geometry car are two common approaches to inspect track geometry. Presently, using accelerations from in-service trains has become a popular track inspection approach, because it is a low-cost way to monitor the rail tracks more frequently. However, due to the noise presented in the collected accelerations, detecting anomalies using manually designed features often results in many false alarms. In this paper, we propose a learning-based anomaly detection approach for monitoring the longitude elevation of track geometry from the dynamic response of an in-service train. We consider the track geometry with a sudden change as an anomaly, measured by the signal energy of slopes of the track geometry. The proposed approach uses a variational autoencoder (VAE) to detect the anomaly. The VAE takes accelerations as input and learns a mapping from the frequency-domain representation of acceleration signals to a low-dimensional latent space that represents the distribution of the observed data. The reconstruction probability, which measures the variability of the distribution of the input data, is used as an anomaly score for indicating how well the input follows the normal pattern. Compared to distance- and density-based anomaly detection methods, such as K-nearest neighbor and clustering, the VAE-based anomaly detection is robust to measurement noise and prevents overfitting because it captures the underlying distribution of the data in a low-dimensional space. Furthermore, the VAE-based method does not require model-specific thresholds for detecting anomalies because it uses a probabilistic measurement instead of reconstruction error as the anomaly score. We validate the proposed VAE-based approach on the vibration dataset from an in-service train. We show that this approach outperforms a baseline model (an autoencoder-based anomaly detection method) in terms of recall, precision, and F1-score. The proposed method also successfully addresses the overfitting problem presenting in recurrent neural network-based methods. The results make the proposed approach a strong candidate for low-cost and frequent track geometry inspection.

In this paper, we characterize the effects of obstructions on footstep-induced floor vibrations to enable obstruction-invariant indoor occupant localization. Occupant localization is important in smart building applications such as smart healthcare and energy management. Maintenance and installment requirements limit the application of current sensing approaches (e.g., mobile-based, RF-based, and pressure-based sensing) in real-life applications. To overcome these limitations, prior work has utilized footstep-induced structural vibrations for occupant localization. The main intuition behind these approaches is that the footstep-induced floor vibration waves take different amounts of time to arrive at different sensors. These Time-Differences-of-Arrival (TDoA) can then be leveraged to locate the footstep by assuming similar velocities between the footstep and various sensor locations. This assumption makes these approaches suitable for open areas; however, real buildings have various types of obstructions (e.g., walls, furniture, etc.) which affect wave propagation velocities and hence significantly reduce localization accuracy. Therefore, the prior work requires unobstructed paths between footsteps and sensors for accurate occupant localization, which increases the sensing density requirement and thus, instrumentation and maintenance costs. We have observed that the obstruction mass is one of the key factors in affecting the wave propagation velocity and reducing the localization accuracy. Therefore, to overcome the obstruction challenge, we localize footsteps by considering different velocities between the footsteps and sensors depending on the existence and mass of obstruction on the wave path. Specifically, we (1) detect and estimate the mass of the obstruction by characterizing the wave attenuation rate, (2) use this estimated mass to find the propagation velocities for localization by modeling the velocity-mass relationship through the lamb wave characteristics, and (3) introduce a non-isotropic multilateration approach which robustly leverages these propagation velocities to locate the footsteps (and the occupants). In field experiments, we achieved average localization error of 0.61 meters, which is (1) the same as the average localization error when there is no obstruction and (2) 1.6X improvement compared to the baseline approach.

Automated structural damage diagnosis after earthquakes is important for improving efficiency of disaster response and city rehabilitation. In conventional data-driven frameworks which use machine learning or statistical models, structural damage diagnosis models are often constructed using supervised learning. Supervised learning requires historical structural response data and corresponding damage states (i.e., labels) for each building to learn the building-specific damage diagnosis model. However, in post-earthquake scenarios, historical data with labels are often not available for many buildings in the affected area. This makes it difficult to construct a damage diagnosis model. Further, directly using the historical data from other buildings to construct a damage diagnosis model for the target building would lead to inaccurate results. This is because each building has unique physical properties and thus unique data distribution. To this end, we introduce a new framework, Physics-Informed Multi-source Domain Adversarial Networks (PhyMDAN), to transfer the model learned from other buildings to diagnose structural damage states in the target building without any labels. This framework is based on an adversarial domain adaptation approach that extracts domain-invariant feature representations of data from different buildings. The feature extraction function is trained in an adversarial way, which ensures that extracted feature distributions are robust to variations of structural properties. The feature extraction function is simultaneously jointly trained with damage prediction function to ensure extracted features being predictive for structural damage states. With extracted domain-invariant feature representations, data distributions become consistent across different buildings. We evaluate our framework on both numerical simulation and field data collected from multiple building structures. The results show up to 90.13 % damage detection accuracy and 84.47 % damage quantification accuracy on simulation data, and up to 100 % damage detection accuracy and 69.93 % 5-class damage quantification accuracy when transferring from numerical simulation data to real-world experimental data, which outperforms the state-of-the-art benchmark methods.

This work addresses the main challenges in real-world application of guided-waves for damage detection of pipelines, namely their complex nature and sensitivity to environmental and operational conditions (EOCs). Different propagation characteristics of the wave modes, their distinctive sensitivities to different types and ranges of EOCs, and to different damage scenarios, make the interpretation of diffuse-field guided-wave signals a challenging task. This paper proposes an unsupervised feature-extraction method for online damage detection of pipelines under varying EOCs. The objective is to simplify diffuse-field guided-wave signals to a sparse subset of the arrivals that contains the majority of the energy carried by the signal. We show that such a subset is less affected by EOCs compared to the complete time-traces of the signals. Moreover, it is shown that the effects of damage on the energy of this subset suppress those of EOCs. A set of signals from the undamaged state of a pipe are used as reference records. The reference dataset is used to extract the aforementioned sparse representation. During the monitoring stage, the sparse subset, representing the undamaged pipe, will not accurately reconstruct the energy of a signal from a damaged pipe. In other words, such a sparse representation of guided-waves is sensitive to occurrence of damage. Therefore, the energy estimation errors are used as damage-sensitive features for damage detection purposes. A diverse set of experimental analyses are conducted to verify the hypotheses of the proposed feature-extraction approach, and to validate the detection performance of the damage-sensitive features. The empirical validation of the proposed method includes (1) detecting a structural abnormality in an aluminum pipe, under varying temperature at different ranges, (2) detecting multiple small damages of different types, at different locations, in a steel pipe, under varying temperature, (3) detecting a structural abnormality in an operating hot-water piping system, under multiple varying EOCs, such as temperature, water flow rate, and inner pressure; and (4) detecting a structural abnormality as the ratio of the damaged pipe's signals in the reference dataset increases.

We propose a multi-task learning approach for estimating both the location and magnitude of damage occurring on an experimental bridge using acceleration signals collected from a passing vehicle. This is a low-cost and low-maintenance indirect structural health monitoring approach in which sensors on the vehicle are used to detect bridge damage. Recently, signal processing and machine learning approaches have been shown to perform well in achieving higher-level structural health monitoring objectives, such as damage localization and quantification. However, these methods not only lack robustness to measurement and model noises, but also require more physical insights. Guided by a theoretical formulation of a simple vehicle-bridge interaction system, our approach preserves the non-linearity of the trend of the acceleration signals as severity changes, and simultaneously localizes and quantifies the damage for minimizing uncertainties propagating from the location estimation. We evaluate our model on an experimental dataset. In the experiments, the damage is represented by a mass with gradually changing magnitude attached at different positions on the bridge. The results show that it can estimate locations of the damage with an accuracy of 0.08 m (3.30% of the total length of the bridge) and changes in severity level with an accuracy of 17.81 grams (8.9% of the maximum severity mass).

Deep learning models have been successfully used in medical image analysis problems but they require a large amount of labeled images to obtain good performance. However, such large labeled datasets are costly to acquire. Active learning techniques can be used to minimize the number of required training labels while maximizing the model's performance. In this work, we propose a novel sampling method that queries the unlabeled examples that maximize the average distance to all training set examples in a learned feature space. We then extend our sampling method to define a better initial training set, without the need for a trained model, by using Oriented FAST and Rotated BRIEF (ORB) feature descriptors. We validate MedAL on 3 medical image datasets and show that our method is robust to different dataset properties. MedAL is also efficient, achieving 80% accuracy on the task of Diabetic Retinopathy detection using only 425 labeled images, corresponding to a 32% reduction in the number of required labeled examples compared to the standard uncertainty sampling technique, and a 40% reduction compared to random sampling.

The surveillance of air pollution is becoming a highly concerned issue for city residents and urban administrators. Fixed air quality stations as well as mobile gas sensors have been deployed for air quality monitoring but with sparse observations over the entire temporal-spatial space. Therefore, an inference algorithm is essential for comprehensive fine-grained air pollution sensing. Conventional physically-based models can hardly be applied to all the scenarios, while pure data-driven methods suffer from sampling bias and overfitting problems. This paper presents a hybrid algorithm for air pollution inference by guiding the data learning process with physical model. The quantitative combination of knowledge from observed dataset and a discretized convective-diffusion model is performed within a multi-task learning scheme. Evaluations show that, benefited from physical guidance, our hybrid method obtains higher extrapolation ability and more robustness, achieving the same performance with 1/8 sample amount and obtaining 31.9% less error in noisy synthesized environment. In a real-world deployment in Tianjin, our algorithm outperforms the pure data-driven model with 9.69% less inference error over a 9-day PM 2.5 data collection.

Individual perspiration level indicates a person&rsquo, s physical status as well as their comfort level. Therefore, continuous perspiration level measurement enables people to monitor these conditions for applications including fitness assessment, athlete physical status monitoring, and patient/elderly care. Prior work on perspiration (sweat) sensing required the user either to be static or to wear the adhesive sensor directly on the skin, which limits users&rsquo, mobility and comfort. In this paper, we present a novel conductive thread-based textile sensor that measures an individual&rsquo, s on-cloth sweat quantity. The sensor consists of three conductive threads. Each conductive thread is surrounded by a braided cotton cover. An additional braided cotton cover is placed outside the three conductive threads, holding them in a position that is stable for measurement. the sensor can be embedded at various locations on a person&rsquo, s clothing. When the person sweats, the cotton braids absorb the sweat and change the conductivity (resistance) between conductive threads. We used a voltage dividing circuit to measure this resistance as the sensor output (DC). We then conducted a sensor calibration to map this measured voltage to the quantity of electrolyte solution (with the same density as sweat) applied to the sensor. We used this sensor to measure individuals&rsquo, perspiration quantity and infer their perceived perspiration levels. The system is able to limit the average prediction error to 0.4 levels when compared to five pre-defined perceived perspiration levels.

Automated monitoring of animal behavior can detect changes in animal welfare and health problems. Different animal behaviors can point to disease, unrest or inadequate management, and detecting such behaviors in real time allows automated monitoring to be a valuable tool in livestock production [2]. When monitoring pigs, important sow events such as oestrus, pregnancy or parturition may also be detected through continuous animal monitoring [1].

We present a method for improving left-right walking gait balance using structural floor vibration sensing by characterizing changes in structural properties in the sensing area. Understanding and measuring human gait balance can be used to assess overall health status, mobility, and rehabilitation progress. The key research challenge is that structural properties in the sensing area may differ from one footstep location to the next, resulting in inaccurate footstep force and balance estimations. To address this challenge, our method performs sensor selection using the insight that some sensors in the sensor network are in a similar structural region as the footstep location and, therefore, are not as effected by the observed variations in structural properties as the other sensors. We evaluate the performance of our method by conducting uncontrolled real-world walking experiments in a residential structure. This evaluation shows that our method achieves a 1.6X reduction in force estimation error and a 2.4X reduction in balance estimation as compared to the baseline approach.

Context-based pairing solutions increase the usability of IoT device pairing by eliminating any human involvement in the pairing process. This is possible by utilizing on-board sensors (with same sensing modalities) to capture a common physical context (e.g., ambient sound via each device's microphone). However, in a smart home scenario, it is impractical to assume that all devices will share a common sensing modality. For example, a motion detector is only equipped with an infrared sensor while Amazon Echo only has microphones. In this paper, we develop a new context-based pairing mechanism called Perceptio that uses time as the common factor across differing sensor types. By focusing on the event timing, rather than the specific event sensor data, Perceptio creates event fingerprints that can be matched across a variety of IoT devices. We propose Perceptio based on the idea that devices co-located within a physically secure boundary (e.g., single family house) can observe more events in common over time, as opposed to devices outside. Devices make use of the observed contextual information to provide entropy for Perceptio's pairing protocol. We design and implement Perceptio, and evaluate its effectiveness as an autonomous secure pairing solution. Our implementation demonstrates the ability to sufficiently distinguish between legitimate devices (placed within the boundary) and attacker devices (placed outside) by imposing a threshold on fingerprint similarity. Perceptio demonstrates an average fingerprint similarity of 94.9% between legitimate devices while even a hypothetical impossibly well-performing attacker yields only 68.9% between itself and a valid device.

In the Internet of Things (IoT) paradigm, it is important to easily setup and control devices, which is achieved by pairing. In this work, we present a novel pairing scheme that utilizes heterogeneous sensing. The core idea is that devices with different sensing capabilities can still detect common information about their user. We demonstrate this idea through an example application consisting of a camera and IoT devices with inertial sensors. As the user holds a device and moves it around, the camera captures the human's pose and compares it to the IoT device motion. If the motion features are similar enough, the device can be successfully paired to the camera's network.

Vulnerability functions often rely on data from expert opinion, post-earthquake investigations, or analytical simulations. Combining the information can be particularly challenging. In this paper, a Bayesian statistical framework is presented to combining disparate information. The framework is illustrated through application to earthquake mortality data obtained from the 2005 Pakistan earthquake and from PAGER. Three different models are tested including an exponential, a combination of Bernoulli and exponential and Bernoulli and gamma fit to model respectively zero and non-zero mortality rates. A novel Bayesian model for the Bernoulli-exponential and Bernoulli-gamma probability densities is introduced. It is found that the exponential distribution represents the zero casualties very poorly. The Bernoulli-exponential and Bernoulli-gamma models capture the data for both the zero and non-zero mortality rates. It is also shown that the Bernoulli-gamma model fits the 2005 Pakistan data the best and has uncertainties that are smaller than either the ones from the 2005 Pakistan data or the PAGER data.

Civil infrastructures are subjected to various loads over their lifetime, which leads to structural degradation. To understand and predict these dynamic behaviors of physical infrastructure...

In this paper, we introduce a method for estimating human left/right walking gait balance using footstep-induced structural vibrations. Understanding human gait balance is an integral component of assessing gait, neurological and musculoskeletal conditions, overall health status, and risk of falls. Existing techniques utilize pressure- sensing mats, wearable devices, and human observation-based assessment by healthcare providers. These existing methods are collectively limited in their operation and deployment; often requiring dense sensor deployment or direct user interaction. To address these limitations, we utilize footstep-induced structural vibration responses. Based on the physical insight that the vibration energy is a function of the force exerted by a footstep, we calculate the vibration signal energy due to a footstep and use it to estimate the footstep force. By comparing the footstep forces while walking, we determine balance. This approach enables non-intrusive gait balance assessment using sparsely deployed sensors. The primary research challenge is that the floor vibration signal energy is also significantly affected by the distance between the footstep location and the vibration sensor; this function is unclear in real-world scenarios and is a mixed function of wave propagation and structure-dependent properties. We overcome this challenge through footstep localization and incorporating structural factors into an analytical force-energy-distance function. This function is estimated through a nonlinear least squares regression analysis. We evaluate the performance of our method with a real-world deployment in a campus building. Our approach estimates footstep forces with a RMSE of 61.0N (8% of participant's body weight), representing a 1.54X improvement over the baseline.

We present Heart and Sole, a shoe-based system that can track a subject's pulse through their feet. Heart rate monitoring can be used to enhance athletic training and assess risk for several health conditions, such as cardiovascular risk and Parkinson's Disease. Many solutions have been proposed to track people's heart rates in their daily lives, but none of them follow the user around without requiring the user to wear special devices. By incorporating sensors into shoes that the user already wears in everyday life, we can create a ubiquitous heart-monitoring system that follows the user around but also remains ambient in the environment. The sensors lie against the foot as part of the tongue of the shoe. The challenge is to differentiate the weak signal induced by arterial motion from the signal noise caused by musculoskeletal movement. Heart and Sole addresses this by amplifying the signal and identifying low motion points where we can more easily extract the signal. Our demo will have a demonstrator wearing a shoe with such a sensor in it, and demonstrating that their measured pulse matches up with the one being measured by our ground truth, a pulse oximeter.

We present a data-driven approach based on physical insights to achieve damage diagnosis of bridges using only vibration signals collected on board the vehicles passing over the bridge. Though data-driven models have been shown to produce promising results in this context, they generally require labeled examples to fit the models (i.e., supervised learning) and make it difficult to interpret the physical mechanisms. We posit that these shortcomings can be alleviated by studying the physical relationship between damage and the distribution of the resulting acceleration signals, and then choosing an appropriate model to invert this process. To help guide the development of appropriate damage diagnosis algorithms, we first make use of the theoretical formulation of the vehicle-bridge interaction system in the frequency domain and conduct a finite element simulation of this system. From the derived numerical solution, we observe that not only is the trend of the acceleration signals of a passing vehicle with different damage severity non-linear, but also that both the low- and high-frequency responses of a passing vehicle contain information about damage severity. Guided by these observations, we use several dimensionality reduction methods to extract representative features from the vehicle’s vibration response. We then propose an unsupervised damage severity comparison model and a semi-supervised damage severity estimation model aiming at indirect monitoring of bridges. We apply the algorithms to diagnose changes that occur in a laboratory bridge model to which a concentrated mass of gradually changing magnitude is attached at mid-span. The experimental results of the damage severity comparison and estimation show that a non-convex and non-linear dimensionality reduction technique (stacked autoencoders) outperforms other linear and/or convex dimensionality reduction techniques. Overall, our results provide evidence for the applicability of indirect structural health monitoring in bridge models and suggest the feasibility of extending this approach to actual structures.

Summary Fragility functions that define the probabilistic relationship between structural damage and ground motion intensity are an integral part of performance-based earthquake engineering or seismic risk analysis. This paper introduces three approaches based on kernel smoothing methods for developing analytical and empirical fragility functions. A kernel assigns a weight to each data that is inversely related to the distance between the data value and the input of the fragility function of interest. The kernel smoothing methods are, therefore, non-parametric forms of data interpolation. These methods enable the implicit treatment of uncertainty in either or both of ground motion intensity and structural damage without making any assumption about the shape of the resulting fragility functions. They are particularly beneficial for sparse, noisy, or non-homogeneous data sets. For illustration purposes, two types of data are considered. The first is a set of numerically simulated responses for a four-story steel moment-resisting frame, and the second is a set of field observations collected after the 2010 Haiti earthquake. The results demonstrate that these methods can develop continuous representations of fragility functions without specifying their functional forms and treat sparse data sets more efficiently than conventional data binning and parametric curve fitting methods. Moreover, various uncertainty analyses are conducted to address the issues of over-fitting, bias, and confidence intervals. Copyright © 2014 John Wiley & Sons, Ltd.

This paper introduces a damage diagnosis algorithm for civil structures that uses a sequential change point detection method. The general change point detection method uses the known pre- and post-damage feature distributions to perform a sequential hypothesis test. In practice, however, the post-damage distribution is unlikely to be known a priori, unless we are looking for a known specific type of damage. Therefore, we introduce an additional algorithm that estimates and updates this distribution as data are collected using the maximum likelihood and the Bayesian methods. We also applied an approximate method to reduce the computation load and memory requirement associated with the estimation. The algorithm is validated using a set of experimental data collected from a four-story steel special moment-resisting frame and multiple sets of simulated data. Various features of different dimensions have been explored, and the algorithm was able to identify damage, particularly when it uses multidimensional damage sensitive features and lower false alarm rates, with a known post-damage feature distribution. For unknown feature distribution cases, the post-damage distribution was consistently estimated and the detection delays were only a few time steps longer than the delays from the general method that assumes we know the post-damage feature distribution. We confirmed that the Bayesian method is particularly efficient in declaring damage with minimal memory requirement, but the maximum likelihood method provides an insightful heuristic approach.

This paper introduces a calibration-free footstep frequency estimation system using footstep-induced structural vibration. Footstep frequency is an important measure for tracking health status in senior/health care and rehabilitation. Using structural vibrations for this estimation can improve intrusiveness commonly associated with long-term monitoring. Because the large number of structure types and the variety of noise they are subjected to, the main challenges of vibration-based approach are: (1) separating footsteps from other impulsive excitations (such as door shutting, cane striking, object droppings, etc.), (2) providing a system which is compatible to different structures and does not require calibration and training for every structure. To combat these challenges, we introduce an online footstep frequency estimation system which uses human walking pattern heuristics to automatically separate and tune the system to distinguish between footstep-induced vibration and other impulsive excitations in different structures. We validate our approach in two different buildings with human participants. The results show that our approach results in F1 score of 0.87, equal to 8× improvement compared to a baseline approach, which classifies the footsteps using a model trained in a different structure.

The number of people passing through different indoor areas is useful in various smart structure applications, including occupancy-based building energy/space management, marketing research, security, etc. Existing approaches to estimate occupant traffic include vision-, sound-, and radio-based (mobile) sensing methods, which have placement limitations (e.g., requirement of line-of-sight, quiet environment, carrying a device all the time). Such limitations make these direct sensing approaches difficult to deploy and maintain. An indirect approach using geophones to measure floor vibration induced by footsteps can be utilized. However, the main challenge lies in distinguishing multiple simultaneous walkers by developing features that can effectively represent the number of mixed signals and characterize the selected features under different traffic conditions. This paper presents a method to monitor multiple persons. Once the vibration signals are obtained, features are extracted to describe the overlapping vibration signals induced by multiple footsteps, which are used for occupancy traffic estimation. In particular, we focus on analysis of the efficiency and limitations of the four selected key features when used for estimating various traffic conditions. We characterize these features with signals collected from controlled impulse load tests as well as from multiple people walking through a real-world sensing area. In our experiments, the system achieves the mean estimation error of ±0.2 people for different occupant traffic conditions (from one to four) using k-nearest neighbor classifier.

The objective of this paper is to characterize frequency-dependent wave propagation of footstep induced floor vibration to improve robustness of vibration-based occupant localization. Occupant localization is an essential part of many smart structure applications (e.g., energy management, patient/customer tracking, etc.). Exist- ing techniques include visual (e.g. cameras and IR sensors), acoustic, RF, and load-based approaches. These approaches have many deployment and operational requirements that limits their adaptation. To overcome these limitations, prior work has utilized footstep-induced vibrations to allow sparse sensor configuration and non-intrusive detection. However, frequency dependent propagation characteristics and low signal-to-noise ratio (SNR) of footstep-induced vibrations change the shape of the signal. Furthermore, estimating the wave propagation velocity for forming the multilateration equations and localizing the footsteps is a challenging task. They, in turn, lead to large errors of localization. In this paper, we present a structural vibration based indoor occupant localization technique using improved time-difference-of-arrival between multiple vibration sensors. In particular we overcome signal distortion by decomposing the signal into frequency components and focusing on high energy components for accurate indoor localization. Such decomposition leverages the frequency-specific propagation characteristics and reduces the effect of low SNR (by choosing the components of highest energy). Furthermore, we develop a velocity calibration method that finds the optimal velocity which minimizes the localization error. We validate our approach through field experiments in a building with human participants. We are able to achieve an average localization error of less than 0.21 meters, which corresponds to a 13X reduction in error when compared to the baseline method using raw data.

SUMMARY Fragility functions are commonly used in performance-based earthquake engineering for predicting the damage state of a structure subjected to an earthquake. This process often involves estimating the structural damage as a function of structural response, such as the story drift ratio and the peak floor absolute acceleration. In this paper, a new framework is proposed to develop fragility functions to be used as a damage classification/prediction method for steel structures based on a wavelet-based damage sensitive feature (DSF). DSFs are often used in structural health monitoring as an indicator of the damage state of the structure, and they are easily estimated from recorded structural responses. The proposed framework for damage classification of steel structures subjected to earthquakes is demonstrated and validated with a set of numerically simulated data for a four-story steel moment-resisting frame designed based on current seismic provisions. It is shown that the damage state of the frame is predicted with less variance using the fragility functions derived from the wavelet-based DSF than it is with fragility functions derived from an alternate acceleration-based measure, the spectral acceleration at the first mode period of the structure. Therefore, the fragility functions derived from the wavelet-based DSF can be used as a probabilistic damage classification model in the field of structural health monitoring and an alternative damage prediction model in the field of performance-based earthquake engineering. Copyright © 2011 John Wiley & Sons, Ltd.

Vehicular crowd sensing systems are designed to achieve large spatio-temporal sensing coverage with low-cost in deployment and maintenance. For example, taxi platforms can be utilized for sensing city-wide air quality. However, the goals of vehicle agents are often inconsistent with the goal of the crowdsourcer. Vehicle agents like taxis prioritize searching for passenger ride requests (defined as task requests), which leads them to gather in busy regions. In contrast, sensing systems often need to sample data over the entire city with a desired distribution (e.g., Uniform distribution, Gaussian Mixture distribution, etc.) to ensure sufficient spatio-temporal information for further analysis. This inconsistency decreases the sensing coverage quality and thus impairs the quality of the collected information. A simple approach to reduce the inconsistency is to greedily incentivize the vehicle agents to different regions. However, incentivization brings challenges, including the heterogeneity of desired target distributions, limited budget to incentivize more vehicle agents, and the high computational complexity of optimizing incentivizing strategies. To this end, we present a vehicular crowd sensing system to efficiently incentivize the vehicle agents to match the sensing distribution of the sampled data to the desired target distribution with a limited budget. To make the system flexible to various desired target distributions, we formulate the incentivizing problem as a new type of non-linear multiple-choice knapsack problem, with the dissimilarity between the collected data distribution and the desired distribution as the objective function. To utilize the budget efficiently, we design a customized incentive by combining monetary incentives and potential task (ride) requests at the destination. Meanwhile, an efficient optimization algorithm, iLOCuS , is presented to plan the incentivizing policy for vehicle agents to decompose the sensing distribution into two distinct levels: time-location level and vehicle level, to approximate the optimal solution iteratively and reduce the dissimilarity objective. Our experimental results based on real-world data show that our system can reduce up to 26.99 percent of the dissimilarity between the sensed and target distributions compared to benchmark methods.

In this paper, the time series based damage detection algorithms developed by Nair, et al. (2006) and Nair and Kiremidjian (2007) are applied to the benchmark experimental data from the National Center for Research on Earthquake Engineering (NCREE) in Taipei, Taiwan. Both acceleration and strain data are analyzed. The data are modeled as autoregressive (AR) processes, and damage sensitive features (DSF) and feature vectors are defined in terms of the first three AR coefficients. In the first algorithm developed by Nair, et al. (2006), hypothesis tests using the t-statistic are applied to evaluate the damaged state. A damage measure (DM) is defined to measure the damage extent. The results show that the DSF\'s from the acceleration data can detect damage while the DSF from the strain data can be used to localize the damage. The DM can be used for damage quantification. In the second algorithm developed by Nair and Kiremidjian (2007) a Gaussian Mixture Model (GMM) is used to model the feature vector, and the Mahalanobis distance is defined to measure damage extent. Additional distance measures are defined and applied in this paper to quantify damage. The results show that damage measures can be used to detect, quantify, and localize the damage for the high intensity and the bidirectional loading cases.

The objective of this paper is to present an occupant detection method through step-induced structural vibration. Occupant detection enables various smart building applications such as space/energy management. Ambient structural vibration monitoring provides a non-intrusive sensing approach to achieve that. The main challenges for structural vibration based occupant footstep detection include that (1) the ambient structural vibration noise may overwhelm the step-induced vibration and (2) there are various other impulse-like excitations that look similar to footstep excitations in the sensing environment (e.g., door closing, chair dragging, etc.), which increase the false alarm rate for occupant detection. To overcome these challenges, a two-stage step-induced signal detection algorithm is developed to (1) incorporate the structural characteristics by selecting the dominant frequencies of the structure to increase the signal-to-noise ratio in the vibration data and thus improve the detection performance and (2) perform footstep classification on detected events to distinguish step-induced floor vibrations from other impulse excitations. The method is validated experimentally in two different buildings with distinct structural properties and noise characteristics, Carnegie Mellon University (CMU) campus building and Vincentian Nursing Home deployments in Pittsburgh, PA. The occupant footstep detection F1 score shows up to 4X reduction in detection error compared to traditional thresholding method.

Ultrasonic guided-waves propagating in pipes with varying environmental and operational conditions (EOCs) are usually the results of complex superposition of multiple modes travelling in multiple paths. Among all of the components forming a complex guided-wave signal, the arrivals scattered by damage (so called scatter signal) are of importance for damage diagnosis purposes. This paper evaluates the potentials of nonlinear decomposition methods for extracting the scatter signal from a multi-modal signal recorded from a pipe under varying temperatures. Current approaches for extracting scatter signal can be categorized as (A) baseline subtraction methods, and (B) linear decomposition methods. In this paper, we first illustrate, experimentally, the challenges for applying these methods on multi-modal signals at varying temperatures. To better analyze the experimental results, the effects of temperature on multi-modal signals are simulated. The simulation results show that different wave modes may have significantly different sensitivities to temperature variations. This brings about challenges such as shape distortion and nonlinear relations between the signals recorded at different temperatures, which prevent the aforementioned methods to be extensible to wide range of temperatures. In this paper, we examine the potential of a nonlinear decomposition method, namely nonlinear principal component analysis (NLPCA), for removing the nonlinear relation between the components of a multi-modal guided-wave signal, and thus, extracting the scatter signal. Ultrasonic pitch-catch measurements from an aluminum pipe segment in a thermally controlled laboratory are used to evaluate the detection performance of the damage-sensitive features extracted by the proposed approach. It is observed that NLPCA can successfully remove nonlinear relations between the signal bases, hence extract scatter signal, for temperature variations up to 10℃, with detection accuracies above 99%.

In spite of their many advantages, real-world application of guided-waves for structural health monitoring (SHM) of pipelines is still quite limited. The challenges can be discussed under three headings: (1) Multiple modes, (2) Multipath reflections, and (3) Sensitivity to environmental and operational conditions (EOCs). These challenges are reviewed in the authors’ previous work. This paper is part of a study whose objective is to overcome these challenges for damage diagnosis of pipes, while addressing the limitations of the current approaches. That is, develop methods that simplify signal while retaining damage information, perform well as EOCs vary, and minimize the use of transducers. In this paper, a supervised method is proposed to extract a sparse subset of the ultrasonic guided-wave signals that contain optimal damage information for detection purposes. That is, a discriminant vector is calculated so that the projections of undamaged and damaged pipes on this vector is separated. In the training stage, data is recorded from intact pipe, and from a pipe with an artificial structural abnormality (to simulate any variation from intact condition). During the monitoring stage, test signals are projected on the discriminant vector, and these projections are used as damage-sensitive features for detection purposes. Being a supervised method, factors such as EOC variations, and difference in the characteristics of the structural abnormality in training and test data, may affect the detection performance. This paper reports the experiments investigating the extent to which the differences in damage size and damage location, as well as temperatures, can influence the discriminatory power of the extracted damage-sensitive features. The results suggest that, for practical ranges of monitoring and damage sizes of interest, the proposed method has low sensitivity to such training factors. High detection performances are obtained for temperature differences up to 14°C. The findings reported in this paper suggest that although the proposed method is a supervised approach, labeling of the training data does not require prior knowledge about the damage characteristics (e.g., size, location). Moreover, the potential of the proposed method for online monitoring is illustrated, for wide range of temperature variations and different damage scenarios.

Multiple ultrasonic guided-wave modes propagating along a pipe travel with different velocities which are themselves a function of frequency. Reflections from the features of the structure (e.g., boundaries, pipe welding, damage, etc.), and their complex superposition, adds to the complexity of guided-waves. Guided-wave based damage diagnosis of pipelines becomes even more challenging when environmental and operational conditions (EOCs) vary (e.g., temperature, flow rate, inner pressure, etc.). These complexities make guided-wave based damage diagnosis of operating pipelines a challenging task. This paper reviews the approaches to-date addressing these challenges, and highlights the preferred characteristics of a method that simplifies guided-wave signals for damage diagnosis purposes. A method is proposed to extract a sparse subset of guided-wave signals in time-domain, while retaining optimal damage information for detection purpose. In this paper, the general concept of this method is proved through an extensive set of experiments. Effects of temperature variation on detection performance of the proposed method, and on discriminatory power of the extracted damage-sensitive features are investigated. The potential of the proposed method for real-time damage detection is illustrated, for wide range of temperature variation scenarios (i.e., temperature difference between training and test data varying between -2°C and 13°C).

Vehicle-based monitoring has the potential to become an accurate and cost-efficient way to monitor infrastructure assets, but a number of challenges must be addressed for such a technique to be implemented widely. The majority of vehicle-based infrastructure sensing has assumed that the vehicle’s speed profile is identical every time it passes over the asset of interest. Ultimately, however this technology will be most practical if damage detection schemes can be applied regardless of the speed of the vehicle. Thus methods must be designed to handle speed variability to make this method more practical. In this paper we investigate the effects of variable speed when monitoring infrastructure from the dynamic response of a passing vehicle, which we measure by placing accelerometers on the vehicle of interest. We have conducted a series of laboratory tests to study this phenomenon, in which a vehicle crosses over a scaled model bridge structure with a varying speed profile. We quantify the ability of several features to detect changes in the infrastructure, independent of the variable speed. We show that aligning signals to normalize for speed variability improves the classification results. This work brings us closer to the ultimate goal of using vehicle-based monitoring to ensure more efficient and more reliable infrastructure in the future.

We present a system for monitoring rail infrastructure from the dynamic response of a train in revenue-service. Presently, rail inspection is performed either visually or with dedicated track vehicles. We propose monitoring the tracks from operational trains, using accelerometers inside the cabin of the train. This would offer an economical approach, as accelerometers are a low-cost technology, tracks would not have to be taken out of service for inspection, and no dedicated inspection vehicles or crews would be required. However, using operational trains presents three challenges: the speed of the train varies from run to run, the position of the train is not known with high accuracy, and the vibration signal recorded in the train’s cabin is filtered by the train and its suspension. We propose extracting features from the signal which are robust to these conditions and show that our proposed features allow for detecting changes in the infrastructure from the train’s dynamic response. We demonstrate our ability to detect subtle track changes both through simulation and using data we collected from Pittsburgh’s Light Rail system. This type of sensing, signal processing, and data analysis could facilitate safer trains and more cost-efficient maintenance in the future. doi: 10.12783/SHM2015/182

In spite of many favorable characteristics of guided-waves for Nondestructive Evaluation (NDE) of pipes, real-world application of these systems is still quite limited. Beside the complexities derived from multi-modal, dispersive and multi-path characteristics of guided-waves, one of the main challenges in guided-wave based NDE of pipelines is sensitivity of these systems to variations of environmental and operational conditions (EOC). This paper investigates the effects of varying EOCs on guilded-wave based NDE of pipelines. We first provide a review of the studies to date in the field of guided-wave based testing to identify research gaps for enhancing the application of these systems in pipeline NDE. To study the identified gaps, guided-wave data from a fully operational piping system, with continuously varying flow rate and temperature, is used. Time-shift and amplitude drift effects due to flow rate variations are evaluated along with those of temperature. It is observed that masking effects of flow rate for damage detection can be at least as significant as temperature effects, and that such effects become more dominant when flow rate and temperature variations co-occur.