AN INTELLIGENT BIM-BASED AUTOMATED PROGRESS MONITORING SYSTEM USING SELF-NAVIGATING ROBOTS FOR DATA ACQUISITION


Thesis Type: Doctorate

Institution Of The Thesis: Middle East Technical University, Faculty of Engineering, Department of Civil Engineering, Turkey

Approval Date: 2018

Thesis Language: English

Student: MUHAMMAD USMAN HASSAN

Supervisor: Aslı Akçamete Güngör

Abstract:

Construction managers require a continuous flow of timely and accurate site status information that is acquired efficiently for successful project delivery. Current methods of data acquisition from the site are error-prone, laborious, and unable to provide timely information to project stakeholders for effective decision making. In this research, we developed a methodology for extraction of data points using BIM, acquisition of progress data using self-navigating robots, estimation of progress information using computer vision algorithms, followed by calculation and visualization of cost metrics. All these steps are performed without any human input in an automated manner to create a robust and efficient mechanism that is both accurate and cost-effective. The developed methodology is named Context-Aware Progress Monitoring System (CAPMS) which consists of five distinct phases. In the first phase; as-built spatial and semantic information from BIM is extracted to calculate data points for element level data acquisition using the imaging sensor. Using this extracted element data, an algorithm creates an element-wise activity list for the formation of a 4D model. The second phase involves acquiring images using a BIM-based data acquisition device, which is verified by a robot, that navigates inside the structures and reaches elements to photograph them. The robot acquires images of building element and transmits them vi to the server for progress estimation from image data. In the third phase, a contextaware method is developed to estimate element status using computer vision algorithms. Contextual information obtained from schedule adds robustness to the developed methodology by reducing false positives. The states of the elements are used to estimate progress status and update cost-based progress metrics which we visualize on a dashboard in the fifth and final phase. The developed system has been validated by using the images obtained on two different construction sites with a robot and processing those images to determine accurate progress status in an automated manner.