Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor | Department of Building and Real Estate | en_US |
| dc.contributor.advisor | Lu, Ming (BRE) | en_US |
| dc.creator | Qi, Kai | - |
| dc.identifier.uri | https://theses.lib.polyu.edu.hk/handle/200/14641 | - |
| dc.language | English | en_US |
| dc.publisher | Hong Kong Polytechnic University | en_US |
| dc.rights | All rights reserved | en_US |
| dc.title | Advancing productivity measurement and resource scheduling through posture-hour based construction planning | en_US |
| dcterms.abstract | The construction industry faces a persistent paradox: while technological advances have revolutionized how engineers design and build structures, the fundamental ways of understanding and managing human work on construction sites remain essentially unchanged, a disconnect that contributes directly to the industry's growing labor shortage. A worker spending an hour bent over tasks, such as rebar tying or concrete finishing, endures significantly higher physical strain than one performing standing tasks, like operating a concrete vibrator or pouring concrete. As construction sites become increasingly instrumented with sensors and cameras, and as computational methods grow more sophisticated, an opportunity emerges to fundamentally reimagine how we conceptualize, measure, and optimize human work in construction. This research develops a framework that transforms construction labor productivity and workforce management from time-based to posture-based planning, leveraging computer vision to quantify physical workload. It introduces scheduling methods that enable practitioners to explicitly consider ergonomic constraints alongside traditional project objectives in terms of time, cost, and quality. | en_US |
| dcterms.abstract | To address this challenge, it is first essential to understand the current state of construction labor productivity (CLP) research and identify gaps in productivity measurement and workforce management knowledge. This comprehensive mapping exposes a critical gap: the limited integration of worker occupational health and safety (OHS) considerations within the productivity framework. | en_US |
| dcterms.abstract | Building on this identified gap, the second contribution of this research develops a novel labor productivity measurement framework that quantifies construction work through workers' physical postures, rather than relying solely on time. The proposed recognition algorithm, which integrates Convolutional Neural Networks (CNN), Bidirectional Long Short-Term Memory (Bi-LSTM), and Multi-Head Attention, achieves a posture classification accuracy of 85.3%. This performance was attained using YOLO-Pose to extract human body keypoints from laboratory-recorded videos and training the model on a dataset partitioned into training, testing, and validation sets. This posture-hour framework provides the granular, ergonomically-informed data necessary to bridge the gap between productivity assessment and worker OHS protection. | en_US |
| dcterms.abstract | With posture-hour data quantifiable through computer vision, the final and most significant contribution of this research is the development of resource-constrained project scheduling methods that integrate these ergonomic insights into construction planning. This research introduces a posture-hour work scheduling framework that incorporates ergonomic constraints directly into construction planning, treating worker posture capacities as schedulable resources alongside traditional considerations, such as crew availability and task dependencies. Validation through real-world concrete-pouring cases demonstrates that while accounting for ergonomic constraints may impact project timelines, it provides policymakers with flexible options to proactively mitigate occupational health risks during the planning stage. Future research will focus on identifying the most reasonable task assignment schemes that can integrate ergonomic safety without increasing the total project duration. Rather than being a burden, this research offers a flexible solution for decision-makers by treating "posture-hours" as a schedulable resource. For instance, ergonomic constraints can be applied as optional parameters to non-critical activities with sufficient schedule slack, improving worker well-being without delaying the project's critical path. | en_US |
| dcterms.abstract | This thesis demonstrates that the long-standing trade-off between productivity and worker occupational health and safety is not inevitable but rather a consequence of lacking adequate metrics and planning methods. Specifically, this research addresses key risk factors for work-related musculoskeletal disorders (WMSD) by quantifying overexertion from high-strain activities, the repetition of physical movement sequences, and the impact of awkward postures—such as bending, squatting, and prolonged arm-raising—that are often overlooked in traditional man-hour-based planning. By establishing posture hours as a measurable, plannable resource, this work provides both researchers and practitioners with a pathway toward truly sustainable construction practices, in which worker well-being becomes an integral part of project planning rather than an afterthought, making the industry more appealing and attractive to new generations. Moreover, by decomposing construction work into quantifiable posture sequences and durations, this research establishes a foundational framework for modeling and programming humanoid robots. Given that humanoid robotics involves high degrees of freedom requiring intricate posture programming and detailed task allocation, this study provides a crucial technical reference for optimizing task dispatching and operational procedures. Specifically, by identifying high-strain patterns and establishing 'posture-hour' thresholds, it offers a strategic pathway for managing the wear-and-tear of robotic joints and components by preventing mechanical overexertion. | en_US |
| dcterms.extent | xvi, 209 pages : color illustrations | en_US |
| dcterms.isPartOf | PolyU Electronic Theses | en_US |
| dcterms.issued | 2026 | en_US |
| dcterms.educationalLevel | Ph.D. | en_US |
| dcterms.educationalLevel | All Doctorate | en_US |
| dcterms.accessRights | open access | en_US |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14641