營建生產作業行為及累積性職業傷害整合自動辨識系統

Motion-Sensing Identification System for Construction Operation and Cumulative Occupational Hazard

楊廷曄(Yang﹐Ting-Yeh)

指導教授 ： 曾仁杰

國立交通大學/工學院/土木工程系所/碩士(2016年)

摘要

在營建管理中，對於生產力的評估，取決於營建工人的單位成本效益，即營建工人的基礎生產力。在營建產業相關研究中，對於基礎生產力的評估有許多不同的方法，如傳統的現場工作抽樣、工地效率評估及短時效率評估法、近年來常用的DEA資料包絡分析與Thomas基準分析法，各種方法的主要目的是評估基準生產力效率值，評估生產力的差異，或確認生產力的變異情況，皆不是針對生產過程作量測。若能利用電腦以自動化的方式，獲得即時性的生產作業行為辨識結果，將可取得客觀量化的生產力評估基準。
有鑒於近年來影像感應技術的進步，透過光學體感器即可進行影像動態追蹤，體感器可藉由拍攝人體之動作，追蹤人體關節骨架，並將其擴展至人因作業領域，應用於生產作業行為辨識中。
本研究承接團隊中另一位成員，何家瑋(2015)的累積性職業傷害評估系統，在系統演算法規則作延伸，建立生產作業行為辨識系統，針對營造業較常見之施工作業，如鋼筋綁紮、模板組立、讀圖溝通、搬運作業、砌磚作業、磁磚鋪貼，以實驗的方式利用體感器取得數據，分析其辨識率及準確率的結果，並將其與累積性職業傷害評估系統整合，建立自動化動作捕捉系統。

關鍵字

體感感測器 ； 營建自動化 ； 生產作業評估 ； 生產力 ； 體感姿態分析

並列摘要

In construction management, we use per unit cost-effective, the based productivity of construction workers to evaluate the productivity. There are several evaluation methods to assess the based productivity in construction industry-related research. For examples, the traditional field sampling, construction site efficiency assessment and short-time efficiency assessment method, DEA data envelopment analysis and benchmark analysis Thomas which are popular recently. The main purpose of all the methods above are for assessment of efficiency values of the reference productivity and the differences in productivity, and confirm the variation of productivity. Not for the measurement of production process. If we can use an automated way by computer to get the recognition result of production operations behavior instantly, we will be able to obtain objective and quantitative assessment of productivity benchmarks.
In view of the recent advances in image sensor technology, the dynamic images can be traced through the optical sensor body. For instance, the optical sensor body can track the body's joints and skeleton action by capturing the images. And then extend this to operations identified in production behaviors.
This study is to undertake the Evaluation System for Cumulative Occupational Hazard of Posture from another member of our team, Chia-Wei Ho (2015). We extend the rules of system algorithms to build Identification System for Construction Operation, which is helpful to common construction work of construction industry. Such as rebar installation, Installation and removal of formworks , image communication, transport operation, bricklaying operation, and tile paving. We use optical sensor body to obtain experimental data, and analyze the results of its recognition rate and accuracy rate. Then integrate this with Evaluation System for Cumulative Occupational Hazard of Posture to create an automated motion capture system.

並列關鍵字

Kinect ； Construction Automation ； Productivity ； Posture Analysis ； Motion-Sensing

參考文獻

[30] 張紹文，(2009)，「應用人體動作分析於數位人動作之產生」， 國立清華大學工業工程與工程管理研究所，碩士論文。

[42] 黃弘銘，(2008)，「改進分析營建基準生產力的方法」，國立高雄第一科技大學工程科技研究所，博士論文

[3] Ibbs, William, Nguyen, Long D., and Lee, Seulkee, (2007), “Quantified impacts of project change.” ASCE Journal of Professional Issues in Engineering Education and Practice. Volume 133, Pages 45-52.

[8] Winch, Graham and Carr, Brid, (2001), “Benchmarking on-site productivity in France and the UK: a CALIBRE approach”, Construction Management and Economics, 19-6, Pages 577 –590.

[9] Hsieh, Ting-Ya, (1998), “Impact of Subcontracting on Site Productivity: Lessons Learned in Taiwan,” Journal of Construction Engineering and Management, Volume 124, No.2.