Korean Journal of Construction Engineering and Management (한국건설관리학회논문집)

Korea Institute of Construction Engineering and Management (한국건설관리학회)
권호 표지
DOI QR코드

DOI QR Code

The Development of Productivity Prediction Model for Interior Finishes of Apartment using Deep Learning Techniques

Deep Learning 기반 공동주택 마감공사 단위작업별 생산성 예측모델 개발 - 내장공사를 중심으로 -

  • Lee, Giryun (Department of Architectural Engineering, Kyung Hee University) ;
  • Han, Choong-Hee (Department of Architectural Engineering, Kyung Hee University) ;
  • Lee, Junbok (Department of Architectural Engineering, Kyung Hee University)
  • Received : 2018.09.20
  • Accepted : 2018.12.27
  • Published : 2019.03.31
Download PDF
⟨ Previous Next ⟩

Abstract

Despite the importance and function of productivity information, in the Korean construction industry, the method of collecting and analyzing productivity data has not been organized. Also, in most cases, productivity management is reliant on the experience and intuitions of field managers, and productivity data are rarely being utilized in planning and management. Accordingly, this study intends to develop a prediction model for interior finishes of apartment using deep learning techniques, so as to provide a foundation for analyzing the productivity impacting factors and predicting productivity. The result of the study, productivity prediction model for interior finishes of apartment using deep learning techniques, can be a basic module of apartment project management system by applying deep learning to reliable productivity data and developing as data is accumulated in the future. It can also be used in project engineering processes such as estimating work, calculating work days for process planning, and calculating input labor based on productivity data from similar projects in the past. Further, when productivity diverging from predicted productivity is discovered during construction, it is expected that it will be possible to analyze the cause(s) thereof and implement prompt response and preventive measures.

국내 건설산업에서 생산성 정보는 중요성과 그 기능에도 불구하고 생산성 데이터의 수집 및 분석 방법이 체계화되어 있지 못하다. 또한 생산성 관리는 대부분 현장관리자의 경험과 직관에 의존하고 있으며 생산성 데이터를 공사계획 및 관리에 적극 활용하지 못하고 있는 상황이다. 따라서 본 연구에서는 공동주택 마감공사의 생산성 예측 및 생산성 영향요인을 분석할 수 있는 기반을 마련하기 위해 단위작업별 생산성 관련 데이터를 수집하여 딥러닝 기반의 생산성 예측모델을 개발하고자 한다. 연구결과인 딥러닝 기반의 공동주택 단위작업별 생산성 예측모델은 신뢰할 수 있는 생산성 정보 데이터에 딥러닝을 적용하여 향후 데이터가 축적될수록 발전되는 기술로 공동주택 프로젝트 관리시스템의 기본 모듈이 될 수 있다. 또한 과거 유사한 프로젝트의 생산성 데이터를 통한 개산견적, 공정계획을 위한 작업일수 산정, 투입인원 산정 등과 같은 프로젝트 엔지니어링 과정에 활용 가능하며 공사 진행 중 예측과 다른 생산성 발견 시 원인 분석에 용이하여 신속한 대응 및 향후 예방이 가능할 것으로 기대된다.

Keywords

GGRHC4_2019_v20n2_3_f0001.png 이미지

Fig. 1. Process of the study

GGRHC4_2019_v20n2_3_f0002.png 이미지

Fig. 2. Deep neural network diagram (Park, 2017)

GGRHC4_2019_v20n2_3_f0003.png 이미지

Fig. 3. Deep learning-based productivity calculation process

GGRHC4_2019_v20n2_3_f0004.png 이미지

Fig. 4. Deep learning-based productivity prediction process

GGRHC4_2019_v20n2_3_f0005.png 이미지

Fig. 5. Deep learning model

GGRHC4_2019_v20n2_3_f0006.png 이미지

Fig. 6. Example of five-minute ratings

GGRHC4_2019_v20n2_3_f0007.png 이미지

Fig. 7. Comparing input labor calculation process (AS-IS, TO-BE)

Table 1. Preliminary study on productivity

GGRHC4_2019_v20n2_3_t0001.png 이미지

Table 2. Productivity-impacting factors in interior works

GGRHC4_2019_v20n2_3_t0002.png 이미지

Table 3. Productivity impacting factor variation rules

GGRHC4_2019_v20n2_3_t0003.png 이미지

Table 4. Deep learning model used functions

GGRHC4_2019_v20n2_3_t0004.png 이미지

Table 5. Overview of gathering data from sample project

GGRHC4_2019_v20n2_3_t0005.png 이미지

Table 6. Baseline and actual productivity of sample project

GGRHC4_2019_v20n2_3_t0006.png 이미지

Table 7. Criteria for classification

GGRHC4_2019_v20n2_3_t0007.png 이미지

Table 8. Five-minute ratings design of sample project

GGRHC4_2019_v20n2_3_t0008.png 이미지

Table 9. Effective work rate and productivity by crew

GGRHC4_2019_v20n2_3_t0009.png 이미지

Table 10. Deep learning model data of the sample project

GGRHC4_2019_v20n2_3_t0010.png 이미지

Table 11. Impact analysis of sample project impacting factor

GGRHC4_2019_v20n2_3_t0011.png 이미지

Table 12. Productivity prediction comparison (AS-IS, TO-BE)

GGRHC4_2019_v20n2_3_t0012.png 이미지

References

  1. Ha, Y.H. (2002). "A study on the development of construction labor management system using information technology" Master's thesis, INHA University.
  2. Huh, Y.K., Lee, J.Y., Yun, S.H., Tae, Y.H., and Ahn, B.R. (2009). "Productivity Measurement and Analysis on Factors in Steel Erection" Journal of the Korea Institute of Building Construction, KIC, 9(2), pp. 31-38. https://doi.org/10.5345/JKIC.2009.9.2.031
  3. Jeon, S.H., and Koo, K.J. (2015). "Comparison between Labor Inputs by Quantity per Unit Method and by Actual Data Method in the Apartment Housing Construction Work" Korean Journal of Construction Engineering and Management, KICEM, 16(1), pp. 110-118. https://doi.org/10.6106/KJCEM.2015.16.1.110
  4. Kim, T.H. (2009). "Analysis of Labor Productivity in Reinforced Concrete Work" Master's thesis, HANYANG University.
  5. Kim, Y.J., Han, J.Y., Shin, D.W., Kim, K.R., Kim, C.D., and Seo, S.W. (2003). "A Tact Planning and Scheduling Process Model for Reduction of Finishing Work Duration in Building Construction Projects" Journal of the Architectural Institute of Korea, AIK, 19(1), pp. 161-168.
  6. Lee, S.H., Kim, Y.M., Kim, J.H., and Kim, J.J. (2011). "Analysis of Delay Factors Based on Importance of Construction Subject-classified in Apartment Finishing Works" Journal of the Korea Institute of Building Construction, KIC, 11(1), pp. 73-81. https://doi.org/10.5345/JKIC.2011.02.1.073
  7. Lim, J.I., Kim, Y.S., Kim, Y.S., and Kim, S.B. (2008). "A Process of Selecting Productivity Influencing Factors For Forecasting Construction Productivity" Korean Journal of Construction Engineering and Management, KICEM, 9(4), pp. 92-100.
  8. Oh, S.W., Kim, M.H., and Kim, Y.S. (2006). "The Application of Data Warehouse for Developing Construction Productivity Management System" Korean Journal of Construction Engineering and Management, KICEM, 7(2), pp. 127-137.
  9. Park, B.R., Choi, E.J., Lee, H.E., Kim, T.W., and Moon, J.W. (2017). "Research Trends for the Deep Learning-based Metabolic Rate Calculation" KIEAE Journal, KIEAE, 17(5), pp. 95-100. https://doi.org/10.12813/kieae.2017.17.5.095
  10. Son, C.B., and Lee, D.C. (2005). "An Analysis on the Selection and Application of Productivity Improving Factors in Apartment Building Construction" Journal of the Architectural Institute of Korea, AIK, 21(4), pp. 133-140.
  11. Woo, G.B., Oh, S.W., Kim, Y.S., and Kim, Y.S. (2008). "The Development of a Productivity Prediction System in the Structural Framework of Apartment Housing Projects Using Data Mining Technique" Journal of the Architectural Institute of Korea, AIK, 24(9), pp. 113-122.