ZEMCH 2019 International Conference Proceedings April.2020 | Page 372
3.2. 4D BIM Model Update Based on Achieved Information
Then, we considered the matter carefully how to update the as‐planned BIM model using
information obtained through the previous process. After acquiring the BIM model of the targeted
building where the photograph was taken, the project coordinates was matched with the actual location
coordinates, and the project direction was also converted to the true north. The story, which is difficult
to grasp from the digital image, was set directly by the worker as the fifth floor. The latitude and
longitude values obtained from the digital photographs were converted to TM coordinates, the drywall
at the corresponding position could be selected, and the comment in the properties window of the BIM
model was set to “in progress”. The wall in the picture was identified in the BIM model by the comment
and the color table set in the method section, and then converted to the current construction state
reflected model as shown in Figure 4.
(a) (b)
Figure 4. Update of 4D BIM by matching location information between digital image and BIM: (a)
Converting position information as TM coordinates and determining the drywall of the location, then
changing its properties as “in‐progress”; (b) The color‐changed wall as “in‐progress” color of yellow.
4. Discussion & Conclusions
In this paper, we proposed a progress monitoring method using site photos for construction
projects.
For this, the authors proposed method with three integrated technologies, and the validity of it
was confirmed as a result of the experimental application. As results of experiment, the development
of S‐CNN model using computer vision outputs of WBS and work status for newly input photographs
shows the high accuracy rates. The proposal of applying the computer vision, machine learning, and
digital GPS position coordinates, the process of updating BIM data immediately, is meaningful
especially not only for its feasibility but also for its novelty. In particular, an existing progress
monitoring work was an additional work of a worker’s daily work, and is likely to cause mistakes such
as omission and miswriting due to the worker’s tiredness. By presenting a method that can easily
manage a process simply with updating after photo taking, this study suggests a very supportive
framework for the actual construction site and the management system from the headquarter office as
well.
However, still there are remained technical limitations such as the way of level positioning, which
is currently set up manually by a worker. Also the proposed progress monitoring method have not
achieved or suggested a platform that can connect these three respective methods and technologies.
Therefore, future research should be extended to propose and commercialize a physical platform the
integrates individually structured methods and systems as one. At the same time, the positioning
tracking system and the location based system with the smartphone are expected to improve the
361
ZEMCH 2019 International Conference l Seoul, Korea