Explore the vast range of titles available.

Ground penetrating radar (GPR) has been widely utilized for non-destructive inspection of civil infrastructure systems such as bridges and tunnels. However, the identification of GPR signatures poses significant challenges due to the overlapped multiple objects. To overcome the obstacle, we proposed an innovative Mask R-CNN based network considering spatial relationship between GPR signatures. Firstly, to capture the spatial relationship of overlapping signatures, we introduced an improved intersection over union considering central distance and aspect ratio between GPR signatures. Secondly, we further modified the Non-Maximum Suppression and enhanced the corresponding anchor generative mechanism. To validate the proposed method, we conducted testing on GPR scans obtained from real data from a bridge. The results demonstrate that the proposed method not only accurately detects GPR signatures, but also significantly outperforms existing Mask R-CNN in terms of segmenting overlapped GPR signature. Specially, the proposed method achieved an average accuracy of 46.8% in the segmentation task, marking a substantial advancement in the field.

This paper presents an ensemble learning approach to predict the active power losses during the allocation and sizing of distributed generation (DG) units in power distribution networks. The forecast model incorporates the Gradient Boosting Machine Regression (GBMR) to estimate DG location, bus voltages, DG size, and active losses without conventional power flow calculations. The results demonstrate that the suggested estimations of power losses and DG sizing are effective, practical, and adaptable for power system management. The accuracy of the proposed model has been validated using key performance metrics and tested on the standard IEEE 33 bus system. In the case of fixed load, the GBMR outperforms other machine learning techniques with the R-squared 0.9997, with a very low mean absolute percentage error (MAPE) (0.2216%) and a root mean square error (RMSE) of 1.0673 in predicting active power losses. This approach is promising in enabling grid operators to effectively manage DG unit integration of distributed energy resources from precise and reliable estimates of the power loss.

The fusion of low-cost sensor networks with air quality stations has become prominent, offering a cost-effective approach to gathering fine-scaled spatial data. However, effective integration of diverse data sources while maintaining reliable information remains challenging. This paper presents an extended clustering method based on the Girvan-Newman algorithm to identify spatially correlated clusters of sensors and nearby observatories. The proposed approach enables localized monitoring within each cluster by partitioning the network into communities, optimizing resource allocation and reducing redundancy. Through our simulations with real-world data collected from the state-run air quality monitoring stations and the low-cost sensor network in Sydney’s suburbs, we demonstrate the effectiveness of this approach in enhancing localized monitoring compared to other clustering methods, namely K-Means Clustering, Density-Based Spatial Clustering of Applications with Noise (DBSCAN) and Agglomerative Clustering. Experimental results illustrate the potential for this method to facilitate comprehensive and high-resolution air quality monitoring systems, advocating the advantages of integrating low-cost sensor networks with conventional monitoring infrastructure.

Deep-learning has emerged as a powerful approach to significantly improve forecast accuracy for air quality estimation. Several models have been developed, demonstrating their own merits in some scenarios and for certain pollutants. In nowcasting, the prediction of air pollution over a small time period essentially demands accurate and reliable estimates, especially in the event cases. From these, selecting the most suitable model to achieve the required forecast performance remains challenging. This paper presents an ensemble framework based on the Dempster-Shafer theory for data fusion to identify the most accurate and reliable forecasts of air pollution obtained from multiple deep neural network models. Our framework is evaluated against three popular machine learning methods, namely, LightGBM, Random Forest, and XGBoost. Experiments are conducted on two horizons: 6-hour and 12-hour predictions using real-world air quality data collected from state-run monitoring stations and low-cost wireless sensor networks.

This paper introduces a comprehensive approach to monorail bridge inspection utilizing unmanned aerial vehicles (UAVs) and digital twin technology. The autonomous UAV-based inspection design encompasses UAV dynamics, tracking control, path planning, and task execution. A dedicated digital twin platform is developed to facilitate rigorous testing and verification of UAV control, mitigating the necessity for extensive physical testing. Methodology validation is achieved through a combination of simulations and real-world experiments, affirming its efficacy in authentic scenarios and demonstrating the potential for advancing infrastructure inspection practices.

There has been an increasing interest in the application of robotic autonomous systems (RASs) for construction and mining, particularly the use of RAS technologies to respond to the emergent issues for earthmoving equipment operating in volatile environments and for the need of multiplatform cooperation. Researchers and practitioners are in need of techniques and developments to deal with these challenges. To address this topic for earthmoving automation, this paper presents a comprehensive survey of significant contributions and recent advances, as reported in the literature, databases of professional societies, and technical documentation from the Original Equipment Manufacturers (OEM). In dealing with volatile environments, advances in sensing, communication and software, data analytics, as well as self-driving technologies can be made to work reliably and have drastically increased safety. It is envisaged that an automated earthmoving site within this decade will manifest the collaboration of bulldozers, graders, and excavators to undertake ground-based tasks without operators behind the cabin controls; in some cases, the machines will be without cabins. It is worth for relevant small- and medium-sized enterprises developing their products to meet the market demands in this area. The study also discusses on future directions for research and development to provide green solutions to earthmoving.

This Special Issue of the Robotica is on recent advances in field and service robotics with a focus on the use of robotic and autonomous technologies to handle tasks in harsh environments and tasks that involve the multirobot cooperation and human–robot interactions.

Emissions from construction activities, particularly in metropolitan areas, are carefully monitored to prevent health problems and environmental degradation. The data quality of low-cost wireless sensors in construction sites remains a challenge for pollution predictive models due to uncertainties of measurement and volatile environment. In this study, we propose a hybrid model using a Long short-term memory integrated with a Bayesian neural network to infer the probabilistic forecasts of particulate matters (i.e., PM1.0, PM2.5, and PM10) emitted from construction activities. The training data are fused by two sources: (1) our developed low-cost wireless sensor network (LWSN) monitoring at a construction site located in Melrose Park, Sydney, Australia, and (2) air-quality stations (AQSs) in four suburbs nearby that monitoring site. The proposed model (LSTM-BNN) is compared with other deep learning methods, namely Gated recurrent unit (GRU), Bidirectional long short-term memory (BiLSTM) and One-dimension convolution neural network (1D-CNN), commonly used for time-series forecast. The experimental results indicate the outperformance of our model to all benchmark models and display a significant improvement at 56.3%, 27.9% and 37.9% in MAEs forecast for all three types of particles compared to a deterministic LSTM model.

This paper proposes an IoT-enabled network of low-cost sensors that are co-located for construction site monitoring. The network performance enhancement is achieved via its system dependability in terms of improved availability, integrity, reliability, maintainability, security and safety in real-time monitoring of environment parameters. The sensor motes of various sensing modules form a reliable wireless in-situ cluster for gathering on-site information of air temperature, soil moisture, air pressure, humidity, particulate matters (PM), emissions and weather variables. They are useful for the site management, improving safety and effective operation of construction equipment. The components for the development include inexpensive microcontrollers ESP32 embedded with wireless gateway function and energy-efficient motes featuring cost-effective sensors. Here, the adoption of the dependability concept for collocated sensor motes aims to introduce a level of redundancy to allow for improving fault-tolerance and reliability. Extensive field tests have been conducted in different environments. Experimental results as well as statistical analysis are provided to verify the merits of the proposed approach.

This paper presents a real-time control system for surface inspection using multiple unmanned aerial vehicles (UAVs). The UAVs are coordinated in a specific formation to collect data of the inspecting objects. The communication platform for data transmission is based on the Internet of Things (IoT). In the proposed architecture, the UAV formation is established via using the angle-encoded particle swarm optimisation to generate an inspecting path and redistribute it to each UAV where communication links are embedded with an IoT board for network and data processing capabilities. Data collected are transmitted in real time through the network to remote computational units. To detect potential damage or defects, an online image processing technique is proposed and implemented based on histograms. Extensive simulation, experiments and comparisons have been conducted to verify the validity and performance of the proposed system.