Application of Visualization Techniques for Construction Progress Monitoring 12Mani Golparvar Fard and Feniosky Peña-Mora
1 PhD Student, Department of Civil and Environmental Engineering, University of Illinois, Urbana-Champaign, 205 N. Mathews Ave, Urbana, IL 61801; PH (217) 333- 2071; FAX (217) 265-8039; email: firstname.lastname@example.org 2 Professor of Construction Management & Information Technology, Department of Civil and Environmental Engineering, University of Illinois, Urbana-Champaign, 205 N. Mathews Ave, Urbana, IL 61801; PH (217) 244-0187; FAX (217) 265-8039; email: email@example.com
For a successful construction project, as-built progress should be constantly monitored and compared with the as-planned construction progress and real-time corrective actions should be taken in case of observed discrepancies. Current media representing these discrepancies (e.g. charts, graphs and still photos) may not facilitate the communication of progress information clearly and quickly which makes the process time-consuming and distracts decision makers from important task of corrective decision making. A series of conceptual visualization techniques (e.g. augmented reality, color and color gradient) have been recently developed to facilitate the communication of progress information and decision making on corrective actions. In this paper, several semi-automated vision-based approaches are applied to further improve and facilitate these processes. To that end, camera matching and registration of the as-planned model with as-built photograph, analyzing the progress status through a material-based detection technique and generation of an occlusion-free photograph are presented.
Despite all the advances in construction equipment and management techniques, inherent complexity and dynamic nature of construction projects and along with the fact that the construction is carried out in an outdoor environment makes it difficult to maintain the as-planned progress during the actual execution (e.g., frequent errors and changes) (Lee and Peña-Mora, 2006). These situations result in schedule and cost overruns, which are prevalent in the construction industry.
One approach to minimize these frequent errors and changes could be to develop an effective real-time progress monitoring process (Slaughter, 1998). If discrepancies between the as-planned and as-built progress are captured promptly, appropriate control actions could be enforced. Hence, the means of representing these discrepancies is one of the keys to support decision making process for control actions (Lee and Peña-Mora, 2006). This decision making process involves and affects many different A/E/C teams and the owners, who have little knowledge about project situations. A recent study by Golparvar Fard (2006) shows most of the time in A/E/C meetings is spent on only describing and explaining the rationale behind the decision making process and less time is spent on value adding tasks such as evaluating and predicting the effects of a decision on the project and also low effectiveness rates on these decision making tasks are reported. One of the major reasons is the lack of proper mediums to visualize and represent information in these meetings (Lee and Peña-Mora, 2006). Particularly, in the case of construction progress monitoring, usually text information (e.g. description of inspection results), graphs and charts (e.g. progress S curve) and still photographs are still used to monitor the results. However these representations do not clearly and quickly show discrepancies for decision making purposes and require project participants to mentally communicate information for decision making tasks and are not sufficient to intuitively show any discrepancy (Lee and Peña-Mora, 2006). It causes the decision making time to be utilized on non value-added decision-making tasks of describing and explaining progress situation. Particularly, when frequent remote and real-time decision making are required in construction, current formats would be very difficult. As an effort to address this issue, the visualization of construction progress monitoring has been introduced through conceptual approaches and is reported as an effective monitoring process (Lee and Peña-Mora, 2006; Kamat and El-Tawil, 2005). Particularly when the user’s capacity and need to understand complex construction
situations are considered, benefits of visualization techniques are maximized.
To further enhance these benefits and expedite the time-consuming procedure of processing and visualizing construction progress, semi-automated visualization techniques are discussed in this paper. It is assumed that progress still photographs are collected from different angles (no fix camera has been available) and the as- planned models are accessible. Along with that, camera matching and registration of the as-planned model with as-built view, analyzing the progress through a material- based detection technique and generating an occlusion-free augmented reality photograph are discussed. In the section that follows, first conceptual visualization techniques are presented. Following to that, the methodology on how to automate visualization of progress monitoring is introduced and our preliminary results are presented.
Conceptual Visualization Techniques
Recently representing construction progress in Augmented Reality (AR) environment has been conceptually explored by Lee and Peña-Mora (2006). In the suggested approach, an as-planned image obtained from a 3D model is superimposed on an as-built still photograph taken on the construction site. For example, Figure 1 demonstrates the as-planned 3D model of a building superimposed on the
Link Link Link Link
Figure 1. Augmented Reality-based Progress Monitoring (Extended from Lee
and Peña-Mora, 2006)
To demonstrate the idea of progress deviations (delays and advances), color and coloring gradients have been recently proposed (Lee and Peña-Mora, 2006; Song et al., 2005). Colors represent construction elements that are behind and/or ahead of the schedule and color gradient can be used to represent work sequences in different work packages while each color would represent a work package. For example in Figure 2, “green” represents elements ahead of schedule and “red” elements that are
behind schedule. On the basis of the deviations detected, progress performance metrics such as SPI (Schedule Performance Index) and CPI (Cost Performance Index) from the Earned Value Analysis can be quantified. These metrics are conceptually represented in Figure 2. Thus, color and gradients provide rich information which aids understanding of the progress situation in real-time and if deviations between as-
planned and as-built progress are quantified, it can facilitate the decision making process. However, considering the time-consuming and tedious data collection and indexing processes, if automated, these conceptual AR visualization techniques would better support progress monitoring. In the section that follows, a series of semi-automated techniques for visualization are discussed.