Contextual multimodal approach for recognizing concurrent activities of equipment in tunnel construction projects

Automation in Construction, Journal Year: 2023, Volume and Issue: 158, P. 105195 - 105195

Published: Nov. 24, 2023

Language: Английский

---

From raw to refined: Data preprocessing for construction machine learning (ML), deep learning (DL), and reinforcement learning (RL) models

Automation in Construction, Journal Year: 2024, Volume and Issue: 168, P. 105844 - 105844

Published: Oct. 24, 2024

Language: Английский

---

The application of “deep learning” in construction site management: scientometric, thematic and critical analysis

Construction Innovation, Journal Year: 2021, Volume and Issue: 22(3), P. 580 - 603

Published: Dec. 26, 2021

Purpose The digital construction transformation requires using emerging technology such as deep learning to automate implementing tasks. Therefore, this paper aims evaluate the current state of in management tasks enable researchers determine capabilities solutions, well finding research gaps carry out more bridge revealed knowledge and practice gaps. Design/methodology/approach scientometric analysis is conducted for 181 articles assess density publications different topics learning-based applications. After that, a thematic gap are analyze contributions limitations key published each area application. Findings indicates that there four main applications management, namely, automating progress monitoring, safety warning workers, managing equipment, integrating Internet things with automatically collect data from site. refers many successful cases site tasks; however, validations recommended test developed additional required consider practitioners workers perspectives implement existing their daily Practical implications This enables directly find solutions develop workable Accordingly, will be reflected on speeding transformation, which strategy over world. Originality/value To best authors’ knowledge, first its kind adopt structured technique researcher/practitioners either adopting these projects or conducting further extend bridging

Language: Английский

---

Current Status and Future Directions of Deep Learning Applications for Safety Management in Construction

Sustainability, Journal Year: 2021, Volume and Issue: 13(24), P. 13579 - 13579

Published: Dec. 8, 2021

The application of deep learning (DL) for solving construction safety issues has achieved remarkable results in recent years that are superior to traditional methods. However, there is limited literature examining the links between DL and management highlighting contributions studies practice. Thus, this study aims synthesize current status on outline practical challenges future opportunities. A total 66 influential articles were analyzed from a technical aspect, such as convolutional neural networks, recurrent general networks. In context management, three main research directions identified: utilizing behaviors, physical conditions, issues. Overall, applying can resolve important with high reliability; therein CNN-based method behaviors most applied percentages 75% 67%, respectively. Based review findings, opportunities aiming address corresponding limitations proposed: expanding comprehensive dataset, improving restrictions due occlusions, identifying individuals who performed unsafe behaviors. This thus may allow identification key areas where further efforts need be made priority.

Language: Английский

---

Motion Tracking of Daily Living and Physical Activities in Health Care: Systematic Review From Designers’ Perspective

JMIR mhealth and uhealth, Journal Year: 2024, Volume and Issue: 12, P. e46282 - e46282

Published: March 14, 2024

Background Motion tracking technologies serve as crucial links between physical activities and health care insights, facilitating data acquisition essential for analyzing intervening in activity. Yet, systematic methodologies evaluating motion data, especially concerning user activity recognition applications, remain underreported. Objective This study aims to systematically review daily living activities, emphasizing the critical interaction among devices, users, environments from a design perspective, analyze process involved application research. It intends delineate intricacies contexts, focusing on enhancing data’s accuracy applicability monitoring intervention strategies. Methods Using review, this research scrutinized their wellness, examining studies Scopus, Web of Science, EBSCO, PubMed databases. The used actor network theory data-enabled understand complex interplay humans, within these applications. Results Out 1501 initially identified studies, 54 (3.66%) were included in-depth analysis. These articles predominantly accelerometer gyroscope sensors (n=43, 80%) monitor motion, demonstrating strong preference capturing both dynamic static activities. While incorporating portable devices (n=11, 20%) multisensor configurations (n=16, 30%), across body (n=15, 28%) spaces (n=17, 31%) highlights diverse applications diversity reflects application’s alignment with types ranging movements specialized scenarios. results also reveal participant pool, including general public, athletes, groups, focus healthy individuals (n=31, 57%) athletes (n=14, 26%). Despite extensive range, primarily laboratory-based (n=39, 72%) aimed at professional uses, such precise identification joint functionality assessment, emphasizes significant challenge translating findings controlled conditions everyday Conclusions study’s comprehensive investigation technology reveals gap methods collection practical real-world proposes an innovative approach that includes designers process, importance framework. ensures is aligned varied nature Such integration developing are accessible, intuitive, tailored meet needs. By leveraging multidisciplinary combines design, engineering, sciences, opens new pathways usability effectiveness technologies.

Language: Английский

---