Search

Understanding pedestrian flow patterns in urban areas could support the decision‐making for infrastructure planning. By incorporating computer vision techniques into surveillance video processing, human walking trajectories in a wide area could be identified by pedestrian re‐identification (ReID) across multiple cameras. Recent ReID methods mostly use convolutional neural networks equipped with deep learning techniques to extract discriminative human features from images for identity matching. However, they still suffer from realistic challenges such as occlusion and appearance variation. This paper develops a ReID‐based framework for pedestrian trajectory recognition across multiple cameras. Specifically, a generic approach of explainable model design is presented, which intuitively analyzes existing baseline models based on feature visualization. Hence, a new model named OSNet + BDB is developed that extracts discriminative‐and‐distributed features. Additionally, an incremental feature aggregation strategy is designed for more robust identity matching. Our ReID method notably outperforms its baselines by 4% identification F1 accuracy in public benchmarks. Practically, pedestrian flow statistics in a real building are extracted for behavioral modeling. Simulations of several what‐if layouts are then conducted for facility performance evaluation. [ABSTRACT FROM AUTHOR]

Wireless sensor networks (WSN) offer an automated, convenient, and low-cost option for building IoT networks for monitoring the structural health of complex infrastructures. The limited bandwidth of a typical WSN is not favorable for intensive data transfer. Therefore, utilizing edge computing that allows raw data to be processed and compressed at IoT devices enables efficient data transfer to the central terminal for data storage and further analytics. However, edge computing requires an additional microcontroller unit (MCU), which significantly increases the edge device power consumption and thus reduces the device's battery lifespan. In this paper, three strategies are developed for saving the battery power of sensor devices with edge computing: (1) managing the connectivity settings of the wireless protocol settings, (2) managing the operations of the edge computing chip, and (3) adding battery capacity with parallel connections. The first strategy focuses on fine-tuning the trade-off between the WSN data throughput and power consumption. The second strategy develops two modes for power saving in certain computationally relaxing periods: Shutdown mode and Low-power mode. The Shutdown mode suspends the operations of the edge computing chip, allowing the edge computing chip to run for a limited time per day. The Low-power mode allows continuous operation of the edge computing chop at low clock speed, which consumes less current. The third strategy doubles the device lifespan by doubling the battery charge capacity using parallel connections. Combining all three strategies significantly enhances the sensor battery lifespan from three weeks to at least six months, notably reducing the frequency of on-site hardware maintenance while providing smooth data transfer due to edge computing. • Developed edge computing IoT crack detection system using Bluetooth Mesh • Developed strategies to minimize mesh power consumption by modifying connection settings and adding battery capacity • Developed strategies to minimize sensor power consumption by controlling edge computing chip operation • Validated the node battery lifetimes through voltage drop graphs • Demonstrated battery lifetime improvement of at least 227% for sensors and 172% for relays [ABSTRACT FROM AUTHOR]

Construction sites are highly hazardous due to the dynamic interaction between workers and moving equipment, with high fatality rates caused by collision and falling from height, etc. Hence, identifying unsafe behaviors among workers is crucial for enhancing site safety, such as tracking their on-site movement and personal protective equipment (PPE). Vision-based video processing has been actively used to automatically recognize workers and their behaviors on construction sites. However, existing studies mainly monitor workers within a single camera capturing only a small sub-region. As workers typically move around fairly large sites, continuously tracking their movement across multiple cameras would enable more comprehensive behavioral analyses. Hence, this paper proposes a framework for monitoring safety compliance among workers, by combining worker re-identification (ReID) and PPE classification. Deep learning-based approaches are developed to address the challenges for these two tasks respectively. For ReID, a new loss function named similarity loss is designed to encourage deep learning models to learn more discriminative human features, realizing a more robust tracking of individual workers. For classifying PPE statuses, a weighted-class strategy is proposed to mitigate model bias when given imbalanced samples among classes, for improved performance despite limited training samples. By combining the ReID and PPE classification results, a workflow is developed to log any incident of workers not wearing the necessary PPEs. With an actual construction site dataset, the proposed methods improve worker ReID and PPE classification by 4% and 13% accuracies respectively, which will facilitate site video analytics and inspection of site safety compliance among workers. • Designed a similarity loss to learn discriminative human features for worker ReID • Designed a weighted-class strategy for PPE classification with imbalanced samples • Monitored worker PPE statuses with consistent identities across multiple cameras • Developed an incident logging workflow to monitor worker PPE compliance on site • Improved ReID and PPE classification by 4% and 13% accuracies over existing CNNs [ABSTRACT FROM AUTHOR]

• Proposed time-series method for crowd prediction with long-time gap (1 day ahead). • Proposed 2D inputs that exploit prior patterns over past days for improved prediction. • Conducted sensitivity analyses of time gap length and input feature size on prediction accuracy. • The LT2D-method improves different baseline models generally by 22% in accuracy. Crowd prediction is a crucial aspect of modern life with innumerable applications. By predicting future human occupancy in advance, crowd prediction can support the decision-making processes of facility stakeholders, e.g., the campus operator can schedule facility maintenance during the period of lowest pedestrian flow to eliminate any disturbance. Conventional crowd prediction utilizes statistical models and rule-based data mining techniques, which are tedious in data processing and error-prone. Hence, this study formulates crowd prediction into a time-series analysis based on deep learning. Despite its wide adaptability in various research fields, deep learning-based time series analysis is seldom adopted in crowd prediction. There are two major limitations in previous studies: firstly, the prediction accuracy notably degrades with increased prediction length, and secondly only the temporal pattern along a single time dimension is exploited, i.e., the consecutive time steps in the most recent input data. Therefore, a Long-Time Gap Two-Dimensional method, entitled LT2D-method, is proposed to increase the crowd prediction length of with high accuracy. The LT2D-method is composed of two parts, (1) long-time gap prediction, which extends the prediction length to 240 time steps (1 day) with high accuracy, and (2) 2D inputs method, which exploits the prior knowledge from different time dimensions to further improve the prediction accuracy of long-time gap prediction. The proposed LT2D-method can be generally adapted to deep learning models, such as LSTM, BiLSTM, and GRU, to improve the prediction accuracy. By incorporating the proposed LT2D-method into different baseline models, the accuracy is generally improved by around 22%, demonstrating the robustness and generalizability of our method. [ABSTRACT FROM AUTHOR]

• A robust framework is proposed to extract pedestrian flow statistics from videos. • Pedestrian trajectories and attributes are jointly exploited for flow analytics. • A tracking model that integrates human appearance and motion metrics is proposed. • An attribute metric and probation strategy are designed for robust identity mapping. • Our pedestrian tracking model outperforms the baseline one by 12% in accuracy. Analyzing the walking behavior of the public is vital for revealing the need for infrastructure design in a local neighborhood, supporting human-centric urban area development. Traditional walking behavior analysis practices relying on manual on-street surveys to collect pedestrian flow data are labor-intensive and tedious. On the contrary, automated video analytics using surveillance cameras based on computer vision and deep learning techniques appears more effective in generating pedestrian flow statistics. Nevertheless, most existing methods of pedestrian tracking and attribute recognition suffer from several challenging conditions, such as inter-person occlusion and appearance variations, which leads to ambiguous identities and hence inaccurate pedestrian flow statistics. Therefore, this paper proposes a more robust methodology of pedestrian tracking and attribute recognition, facilitating the analysis of pedestrian walking behavior. Specific limitations of a current state-of-the-art method are inferred, based on which several improvement strategies are proposed: 1) incorporating high-level pedestrian attributes to enhance pedestrian tracking, 2) a similarity measure integrating multiple cues for identity matching, and 3) a probation mechanism for more robust identity matching. From our evaluation using two public benchmark datasets, the developed strategies notably enhance the robustness of pedestrian tracking against the challenging conditions mentioned above. Subsequently, the outputs of trajectories and attributes are aggregated into fine-grained pedestrian flow statistics among different pedestrian groups. Overall, our developed framework can support a more comprehensive and reliable decision-making for human-centric planning and design in different urban areas. The framework is also applicable to exploiting pedestrian movement patterns in different scenes for analyses such as urban walkability evaluation. Moreover, the developed mechanisms are generalizable to future researches as a baseline, which provides generic insights of how to fundamentally enhance pedestrian tracking. [ABSTRACT FROM AUTHOR]

Building Information Modeling (BIM) is widely adopted for design collaboration in the architecture, engineering, and construction (AEC) industry, while its centralized paradigm suffers the risk of data manipulation. To this end, blockchain is an emerging distributive technology that guarantees data authenticity and integrity by providing decentralized, immutable, and traceable data storage. However, directly integrating BIM with blockchain risks leaking sensitive data, since a blockchain is a transparent network insofar as shared information (including sensitive data) can be disclosed to all members without access control. Therefore, this paper proposes a confidentiality-minded framework (CMF) for blockchain-based design collaboration. The major innovations are twofold. First, an access control model is developed in the CMF to prevent unauthorized access to sensitive BIM data in a blockchain ledger. Second, new design strategies are developed in CMF to facilitate design coordination within the access-controlled blockchain network. An illustrative design example validates the feasibility and performance of the proposed CMF, which has acceptable latency and storage cost. The results also show that the sensitive BIM data are effectively kept confidential when project members collaborate within the CMF. • Proposed an access control model to protect sensitive BIM data in the blockchain ledger. • Developed BIM component-based workflows for design coordination in an access-controlled blockchain network. • Developed a BIM Merkle tree (BMT) data model and BMT algorithms for efficient versioning of BIM components. • Validated the feasibility and performance of CMF through an illustrative example. • Results show that the CMF can protect BIM confidentiality in the blockchain-based design collaboration. [ABSTRACT FROM AUTHOR]

The variation of construction machine poses is one of the main causes for interactive on-site safety issues such as struck-by hazards. With the aim to reduce such hazards, we propose a framework for predicting construction machine poses based on historical motion data and activity attributes. After building a machine motion dataset, we develop a keypoint-based method for recognizing machine activities considering working patterns and interaction characteristics. The recognized activity information is then incorporated with historical pose data to predict future machine poses through a type of recurrent neural network (RNN), named Gated Recurrent Unit (GRU). In experiments of using excavators as the objects, our framework achieves decent performance for machine pose prediction, which is further improved by incorporating activity information, reaching an average percentage of correct keypoints (PCK) of 90.22%. The results indicate the high potential of our framework in predicting construction machine poses and improving on-site safety. • Proposed a GRU-based method for machine pose prediction using motion capture data • Improved machine pose prediction results by keypoint-based activity recognition • Considered non-geometric data like machine interaction and activity type • A rollback method greatly reduced the influence of uncertainty in activity recognition • Achieved an average accuracy over 90% for machine pose prediction [ABSTRACT FROM AUTHOR]

Construction sites are among the most hazardous places with various safety issues. The high rate of hazards on construction sites can be attributed to the dynamic and complex characteristics of construction-related entities, such as the movement of construction equipment and workers as well as the interactions among them. Tracking construction equipment and workers can help avoid potential collisions and other accidents to achieve safer on-site conditions. As construction equipment (e.g. excavators, trucks, cranes, and bulldozers) plays a significant role in construction projects, it is important to track the location, pose and movement of construction equipment. Currently, with the wide installation of surveillance cameras on construction sites, computer vision techniques are explored to process the captured surveillance videos and images, such as to monitor the site conditions and prevent potential hazards. Previous studies have attempted to identify and locate different types of construction equipment on construction sites based on surveillance videos using computer vision techniques. However, there are limited studies that automatically estimate the full body pose and movement of on-site construction equipment, which can greatly influence the safety condition of construction sites and the utilization of the equipment itself. In this study, a methodology framework is developed for automatically estimating the poses of different construction equipment in videos captured on construction sites using computer vision and deep learning techniques. Firstly, keypoints of equipment are defined, based on which the images collected from the surveillance cameras are annotated to generate the ground truth labels. 70%, 10%, and 20% of the annotated image dataset are used for training, validation and testing, respectively. Then, the architectures of three types of deep learning networks i.e. Stacked Hourglass Network (HG), Cascaded Pyramid Network (CPN), and an ensemble model (HG-CPN) integrating Stacked Hourglass and Cascaded Pyramid Network are constructed and trained in the same training environment. After training, the three models are evaluated on the testing dataset in terms of normalized errors (NE), percentage of correct keypoints (PCK), area under the curve (AUC), detection speed, and training time. The experiment results demonstrate the promising performance of our proposed methodology framework for automatically estimating different full body poses of construction equipment with high accuracy and fast speed. It is indicated by experiments that both HG and CPN can achieve relative high accuracy, with a PCK value of 91.19% and 91.78% respectively for estimating the equipment full body poses. In addition, the ensemble model with online data augmentation can further improve the accuracy, achieving a NE of 14.57 × 10−3, a PCK of 93.43%, and an AUC of 39.72 × 10−3 at the detection speed of 125 millisecond (ms) per image. This study lays the foundation for applying computer vision and deep learning techniques in the full body pose estimation of construction equipment, which can contribute to the real-time safety monitoring on construction sites. • A methodology framework is proposed to estimate on-site equipment full body poses. • Equipment keypoints are defined and an equipment pose dataset is built and labeled. • Two models (HG and CPN) achieved accuracies over 90% for estimating equipment full body pose. • The ensemble model (HG-CPN) further improved the full body pose estimation accuracy. • HG, CPN, and HG-CPN are compared, and insights are provided for potential improvement. [ABSTRACT FROM AUTHOR]

Rapid design and construction of mobile cabin hospitals (MCHs) have become imperative in the COVID-19 response. However, due to unique design specifications (e.g., parallel design and model pre-revision), collaboration in emergency construction projects (ECPs) like MCHs presents data security vulnerabilities, including a lack of traceability and transparency. These hazards invariably reduce design effectiveness, leading to undesirable rework and project delay. Blockchain technology is a potential solution to address the aforementioned security issues in ECPs because it offers immutable and traceable data storage. Nevertheless, directly implementing blockchain in ECPs is impractical, for the blockchain has a complex deployment process and provides limited functions supporting BIM-based design. Therefore, this paper develops a lightweight blockchain-as-a-service (LBaaS) prototype to enhance the ECPs design efficiency by securing and automating information exchange while eliminating the difficulties of deploying and using blockchain. This paper contributes three elements: (1) Security vulnerabilities of design in ECP are identified. Taking an MCH in Hong Kong as an example, this paper investigates its design process and determines two design characteristics and associated security flaws. (2) Key technologies to support easy deployment and usage of blockchain in ECPs are developed. New technical elements, including a Multi-to-One mapping (MtOM) kit for easy blockchain registration, an integrated workflow retaining existing design practices, and smart contracts for secure interaction with blockchain, are developed to support LBaaS functionality. (3) An LBaaS prototype is validated and evaluated. The prototype is illustrated and evaluated using design examples based on actual MCH project data. Results show that the LBaaS is a feasible and secure approach for ECPs collaboration. This paper deepens the understanding of data security issues in ECPs and offers technical guidance in establishing blockchain solutions. • Identified security vulnerabilities of BIM-based collaboration in emergency construction projects (ECPs). • Exploring the mechanism of integrating ECPs design with a lightweight blockchain-as-a-service (LBaaS). • Developed key technical elements to support functionalities of an LBaaS prototype. • Validated LBaaS prototype via three design examples from an actual mobile cabin hospital project. • Results show that the LBaaS is a practical solution for improving data security and ECP design efficiency. [ABSTRACT FROM AUTHOR]

To reduce machine-related accidents on sites, automatically monitoring the full-body poses of operating heavy machines is crucial. Conventional pose estimation systems relying on homogeneous sensors are vulnerable to negative environmental impacts, leading to inaccurate and unstable estimation of machine states. Hence, a full-body pose estimation framework is proposed for excavators, with a data fusion strategy to utilize different types of onboard sensors for enhanced accuracy and robustness. Specifically, a non-invasive onboard visual-inertial sensor system is designed for data fusion. Then, through competitive and complementary data fusion, the keypoints describing the full-body poses of the excavator are tracked in 3D space. Especially, an EKF-based localization algorithm is developed for optimized multi-keypoint tracking, which is verified to improve the accuracy and robustness of pose estimation by a real-world excavator case study. The proposed sensor-fusion method can effectively improve operational safety, by accurately monitoring the motion of heavy machines operating on construction sites. • A pose estimation framework for excavators proposed using visual-inertial sensor fusion method. • A non-invasive onboard multi-sensor system designed for data fusion to track motions. • An EKF-based algorithm developed to track multi-keypoint of excavators competitively. • Improved accuracy and robustness in keypoint tracking, confirmed by real-life excavator case study. [ABSTRACT FROM AUTHOR]

Monitoring unsafe operations of construction machines is important for reducing accidents on sites. To automate safety monitoring, this paper studies pose estimation of construction machines, which describes their motions in 3D space. Inertial Measurement Unit (IMU) is considered accurate, cost-saving and user-friendly for estimating the pose of machines. Nevertheless, the complexity and instability of construction sites hinder the application of IMU to construction machines. Hence, this study first defines the information requirements of IMU-based full-body pose estimation for construction machines. Then, an efficient method is developed to generate the trajectories of a machine based on kinematics. Moreover, an optimal installation scheme for IMU sensors is investigated systematically to optimize the location and number of IMUs to install. Overall, this study contributes an IMU-based method to provide full-body poses of machines without environmental constraints on real construction sites, and a theoretical basis of optimal IMU installation strategy for reliable motion-related analyses. • The information requirements are defined on IMU-based analysis of construction machine poses. • A trajectory generation method is proposed to describe the change of joint angles. • The practicability and performance of the proposed approach are fully verified in real scenarios. • The average RMSE between the estimated pose and ground truth is 1.71 degrees. • An optimal IMU installation strategy is proposed including location and number. [ABSTRACT FROM AUTHOR]