Search

This study developed a methodology that can use real-time strain data for the assessment of scaffolding safety conditions. The researchers identified 23 safety cases of individual member failure with generic global failure for a four-bay, three-story scaffold model and used scaffold member strain values to identify potential failure cases. A computer simulation on the scaffold model generated the strain datasets required for classification with a support vector machine (SVM). The SVM classification demonstrated a stable prediction accuracy after training with a certain number of strain datasets. Furthermore, the 2nd order polynomial kernel function resulted in better prediction compared to other SVM kernel functions. These results imply that the real-time assessment of scaffolding structures is possible with a limited number of training data for machine-learning classification. Key words: construction, scaffold, safety, machine learning, temporary structure. Resume : Cette etude a elabore une methodologie qui peut utiliser les donnees de deformation en temps reel pour l'evaluation des conditions de securite des echafaudages. Les chercheurs ont identifie 23 cas de defaillance d'un element individuel avec defaillance globale generique pour un modele d'echafaudage a trois etages a quatre baies et ils ont utilise les valeurs de deformation des elements de l'echafaudage pour identifier les cas de defaillance possibles. Une simulation par ordinateur sur le modele d'echafaudage a genere les ensembles de donnees de deformation necessaires a la classification a l'aide d'une machine a vecteurs de support (MVS). La classification de la MVS a demontre une precision de prevision stable apres l'entrainement avec un certain nombre d'ensembles de donnees de deformation. De plus, la fonction de noyau polynomial de 2e ordre a permis une meilleure prevision que les autres fonctions de noyau MVS. Ces resultats impliquent que l'evaluation en temps reel des structures d'echafaudage est possible avec un nombre limite de donnees d'entrainement pour la classification de l'apprentissage machine. [Traduit par la Redaction] Mots-cles : construction, echafaudage, securite, apprentissage machine, structure temporaire., Introduction Background The safety of construction workers is the highest priority in the construction industry (Wu et al. 2015). However, the fatalities reported in the construction industry (OSHA 2018a) demonstrate [...]

Past construction research has focused primarily on the realization of tracking systems but much less on their effectiveness. Despite significant achievements, little is known about the relationship between the environmental site conditions and deployed sensors and their influence on system effectiveness. To reduce this knowledge gap, we investigate the effect of cement-based material on signal energy loss using electromagnetic simulation. We take conventional electromagnetic equations and derive equations that quantitatively describe the signal behavior influenced by cement-based material. Then, we conduct high frequency structure simulator (HFSS) simulations and use the experimental measurements for benchmark comparison with our analytical solution. Results suggest that our analytical method accurately quantifies attenuated signal loss. Also, we construct an attenuation spectrum of energy loss for cement-based material, which can serve as a reference for sensor deployment. Therefore, the method should enable quantification of the signal loss of sensors without actually deploying them or measuring signal data, which should lead to the more effective deployment of tracking sensors. Key words: radio signal strength indication (RSSI), electromagnetic simulation, indoor localization, mobile sensing, optimization, sensor deployment, tracking. Dans le passe, la recherche en construction a porte principalement sur la realisation de systemes de controle des signaux, mais beaucoup moins sur leur efficacite. Malgre des accomplissements importants, on en sait tres peu sur la relation entre les conditions environnementales de l'emplacement et les capteurs deployes et leur effet sur l'efficacite du systeme. Dans le but de reduire cet ecart de connaissances, nous examinons l'effet des materiaux a base de ciment sur la perte d'energie du signal, et ce, en utilisant la simulation electromagnetique. Nous prenons des equations electromagnetiques traditionnelles et derivons des equations qui decrivent quantitativement le comportement du signal influence par les materiaux a base de ciment. Ensuite, nous effectuons une simulation a l'aide d'un simulateur de structure a haute frequence (<< HFSS >>) et des mesures experimentales aux fins de comparaison de reference avec notre solution analytique. Les resultats suggerent que notre methode analytique quantifie precisement la perte de signal attenuee. Aussi, nous construisons un spectre d'attenuation de perte d'energie pour les materiaux a base de ciment, qui peut servir de reference pour le deploiement de capteurs. Donc, la methode devrait permettre la quantification de la perte de signal des capteurs sans pour autant les deployer ou mesurer des donnees de signal, ce qui devrait mener au deploiement plus efficace de capteurs de suivi. [Traduit par la Redaction] Mots-cles: indicateur de la force du signal recu (<< RSSI >>), simulation electromagnetique, localisation interieure, captage mobile, optimisation, deploiement de capteurs, suivi., Introduction Innovation in construction Capitalizing on technological advances in recent years, researchers have explored various methods that benefit the construction industry from design to operation and management. As a result, [...]

Effective crane management is imperative to ensure that construction projects are completed on time, within budget, and with a high degree of safety. Prior research on crane management has mostly focused on satisfaction of spatial constraints and minimization of overlapping areas for operation of multiple cranes. However, little attention has been paid to maximizing concurrent crane operations that justify interdependencies among paths, cranes, overlapping areas, and tasks. To overcome this limitation, this study aims to improve multicrane layout planning comprehensively by accounting for various crane operational and managerial parameters (testing various crane specifications and layouts along with alternative cranes, motion paths, and task execution queues) and their parametric interdependencies. Results of the study on an actual construction site demonstrated the improved ability of the developed model in terms of efficiency compared with existing approaches. Therefore, this research anticipates assisting project managers and practitioners with a model for efficient decision making regarding site layout plans for operation of multiple tower cranes. [ABSTRACT FROM AUTHOR]

Crash narratives provide valuable information to understand traffic crashes and develop roadway safety countermeasures. However, manually reading long text narratives is time-consuming and error-prone. This study presents a deep-learning and clustering-based approach to identifying contributors to traffic crash severity in text narratives. We evaluate the approach using a dataset of narratives from Massachusetts and compare different deep-learning models for semantic similarity. The approach clusters semantically similar phrases in the narratives and provides an overview of frequent topics related to severe crashes, offering a valuable tool for roadway safety analysis and countermeasure development. [ABSTRACT FROM AUTHOR]

Previous research has demonstrated a link between prior knowledge and student success in engineering courses. However, while course-to-course relations exist, researchers have paid insufficient attention to internal course performance development. This study aims to address this gap—designed to quantify and thus extract meaningful insights—by examining a fundamental engineering course, Statics, from three perspectives: (1) progressive learning reflected in performance retention throughout the course; (2) critical topics and their influence on students' performance progression; and (3) student active participation as a surrogate measure of progressive learning. By analyzing data collected from 222 students over five semesters, this study draws insights on student in-course progressive learning. The results show that early learning had significant implications in building a foundation in progressive learning throughout the semester. Additionally, insufficient knowledge on certain topics can hinder student learning progression more than others, which eventually leads to course failure. Finally, student participation is a pathway to enhance learning and achieve excellent course performance. The presented analysis approach provides educators with a mechanism for diagnosing and devising strategies to address conceptual lapses for STEM (science, technology, engineering, and mathematics) courses, especially where progressive learning is essential. [ABSTRACT FROM AUTHOR]