Seniors who live alone at home are at risk of falling and injuring themselves and, thus, may need a mobile robot that monitors and recognizes their poses automatically. Even though deep learning methods are actively evolving in this area, they have limitations in estimating poses that are absent or rare in training datasets. For a lightweight approach, an off-the-shelf 2D pose estimation method, a more sophisticated humanoid model, and a fast optimization method are combined to estimate joint angles for 3D pose estimation. As a novel idea, the depth ambiguity problem of 3D pose estimation is solved by adding a loss function deviation of the center of mass from the center of the supporting feet and penalty functions concerning appropriate joint angle rotation range. To verify the proposed pose estimation method, six daily poses were estimated with a mean joint coordinate difference of 0.097 m and an average angle difference per joint of 10.017 degrees. In addition, to confirm practicality, videos of exercise activities and a scene of a person falling were filmed, and the joint angle trajectories were produced as the 3D estimation results. The optimized execution time per frame was measured at 0.033 s on a single-board computer (SBC) without GPU, showing the feasibility of the proposed method as a real-time system.

Major metropolitan cities worldwide have extensively invested to secure utilities and build state-of-the-art infrastructure related to underground fluid transportation. Sewer and water pipelines make our lives extremely convenient when they function appropriately. However, leakages in underground pipe mains causes sinkholes and drinking-water scarcity. Sinkholes are the complex problems stemming from the interaction of leaked water and ground. The aim of this work is to review the existing methods for monitoring leakage in underground pipelines, the sinkholes caused by these leakages, and the viability of wireless sensor networking (WSN) for monitoring leakages and sinkholes. Herein, the authors have discussed the methods based on different objectives and their applicability via various approaches—(1) patent analysis; (2) web-of-science analysis; (3) WSN-based pipeline leakage and sinkhole monitoring. The study shows that the research on sinkholes due to leakages in sewer and water pipelines by using WSN is still in a premature stage and needs extensive investigation and research contributions. Additionally, the authors have suggested prospects for future research by comparing, analyzing, and classifying the reviewed methods. This study advocates collocating WSN, Internet of things, and artificial intelligence with pipeline monitoring methods to resolve the issues of the sinkhole occurrence.

A sinkhole is a ground surface depression that may occur with or without any indications on the surface and often pose danger to both properties and people. Leakage from underground pipe mains in urban areas may cause sudden ground subsidence or sinkholes. For a long time, researchers have been working on the hazard and risk assessment of sinkhole formation, especially natural sinkholes. However, much less work has been done on risk prediction and the mechanism of manmade sinkholes. In this study, different versions of small-scale sinkhole physical models were used in experiments to monitor ground surface settlement or collapse due to leakage from an underground pipeline. The factors under consideration were the type of subsurface soil profile, type of water flow, and leakage position in the pipeline. The ultimate goal was to use this information to predict the risk of sinkhole occurrence due to leakage from sewer or water pipelines under different subsurface soil conditions. The experimental results and statistical analysis showed that the subsurface soil strata conditions dominated the mechanism of sinkhole occurrence, although other factors also have contributed to the settlement. Then, this analysis was used to predict the sinkhole risk level under different conditions. The development of a reliable sinkhole risk prediction system can potentially minimize the risk to human lives and infrastructure. These findings can be applied to the development of a sinkhole risk index (SRI) that considers various other factors influencing sinkhole occurrence.

Ergonomic risk assessment is critical for identifying working posture hazardous to the health of construction workers. Work-related musculoskeletal disorders (WMSDs) are predominant non-fatal injuries in the construction industry owing to manual handling activities and poor working conditions. However, there is a lack of scientific synopsis aiming to better understand the emerging research focus in this field. To fill the research gap, this study performed a scientometric evaluation of the bibliometric data on ergonomic risk assessment from the Web of Science database using VOSviewer software. The purpose of this study is to analyze the co-occurrence network of keywords, co-authorship network, most active countries, and the sources of publication. The results indicate that research related to risk assessment in construction has fluctuating growth, peaking in 2020 with significant advancements in the USA, China, and Canada. WMSDs, risk factors, construction workers, and ergonomics are hot research topics in this field. Furthermore, the research gaps of previous studies and suggestions for future research have been provided to bridge the knowledge gap. We believe that this scientometric review provides useful reference points for early-stage researchers as well as beneficial in-depth information to experienced practitioners and scholars in the construction industry.

Occupational health and safety hazards associated with confined places are mainly caused by exposure to toxic gases and oxygen deficiency. Lack of awareness, inappropriate monitoring, and improper evacuation methods can lead to worker fatalities. Although previous studies have attempted to develop systems to solve this issue, limited research is available on their application in confined places. In this study, a real-time helmet-mounted system was developed to monitor major toxic gases (methane (CH₄), hydrogen sulfide (H₂S), ammonia (NH₃), and carbon monoxide (CO)), oxygen, temperature, and humidity. Workers outside and inside confined spaces receive alerts every second to immediately initiate the rescue operation in the event of a hazard. The test results of a confined environment (wastewater treatment unit) highlighted that concentrations of CH₄ and H₂S were predominant (13 ppm). Compared to normal atmosphere, CH₄ concentration was 122- and 130-fold higher in the landfill and digestion tanks, respectively, while H₂S was 36- and 19-fold higher in the primary and secondary clarifiers, respectively. The oxygen content (18.2%) and humidity (33%) were below the minimum required limits. This study will benefit future research to target appropriate toxic gas monitoring and alert workers by studying the existing issues and associated factors in confined places.