skip to content
 

Construction Automation

Current Projects

TOP-DOWN-AS-IS MODELLING OF INDUSTRIAL FACILITIES October 2016 – September 2020

EPSRC & AVEVA Group Plc Grant, PI: Ioannis Brilakis, £102,536

This research will explore ways to detect existing building objects in spatial and visual data for the purpose of automating the generation of as-built geometric models of industrial facilities. The research will address both construction engineering and computer vision issues. Leading researcher: Eva Agapaki.

 

AUTOMATIC CONSTRUCTION MONITORING THROUGH SEMANTIC INFORMATION MODELLING April 2017 – March 2020

Discovery Grants, CI: Ioannis Brilakis, External PI: Prof. Xiangyu Wang, Curtin University £202,016

 

Past Projects

COMPUTER VISION AUTOMATED PRODUCTIVITY MEASUREMENT October 2013 – August 2017

EPSRC and Laing O’Rourke Ltd., Grant No. 13440016

The objective of this research is to investigate the performance of a new method that can automatically detect the task cycles of construction activities, measure their duration, and compile reliable statistics based on 100% of the observed data. This can be achieved by applying primarily computer vision and machine learning techniques to process jobsites’ video data. The extracted construction entities’ trajectories will then be used for the detection of possible repeated work cycles by implementing statistical pattern recognition models. An overview of the project can be found in the attached figures. Leading researcher: Eirini Konstantinou.

 

RP 11-12 TRAINING AND CERTIFICATION FOR CONSTRUCTION INSPECTORS May 2011 – May 2013

Georgia Department of Transportation (GDOT), PI: Ioannis Brilakis / $200,000

The purpose of this research effort was to provide material for training and certification courses to GDOT construction inspectors as well as a delivery system which helps inspectors learn inspection methods and techniques in an easy to understand, comprehensible fashion. The project also aimed to improve construction inspection practices employed by the GDOT.

Under this research effort, existing literature regarding inspection practices was reviewed to identify best practices for each type of construction inspection. These practices were then used to develop the manual for each construction inspection. The final manual was then used to convert material into training and certification modules for GDOT inspectors. Leading researchers: Linda Hui, Stefania Radopoulou and Eric Marks.

 

OPRICAL SURVEYOR January 2012 – December 2012

PI: Ioannis Brilakis, $50,000

Detailed measurements are required at each step of the construction process to ensure the new installation, such as a roof or new countertop, fits. These are currently acquired either by hand measuring on the low end or using laser based devices on the high end. A machine vision-based technology is developed to facilitate this process and significantly reduce the cost of operation. The goal of this project was to commercialize this technology for creating 3D drawings of a structure or space. Leading researcher: Habib Fathi.

 

PROGRESSIVE SITE MODELING WITH VIDEOGRAMMETRY August 2008 – July 2011

NSF Grant #0800170, PI: Ioannis Brilakis, $231,407

The objective of this research was to test the hypothesis that a mobile, calibrated set of high resolution video cameras can be used to acquire the spatial data of a construction site with the assistance of a novel videogrammetric method. Under this method, video streams are initially collected from the camera set that is progressively traversed around a construction site. The possible correspondences of each point in each video camera's view are computed (epipolar lines) and the corresponding points are matched using a novel window similarity matching method that compares the video frame along each epipolar line. Based on each match and the camera calibration, the depth value of each point is computed, and the depth map (point cloud) of the scene is generated. In each subsequent frame, all points with a previously identified correspondence in the other video camera's frame are tracked using established 2D point tracking techniques. The resulting point cloud at each frame is then converted to a 3D surface using intelligent proximity algorithms, and the visual data are overlaid to produce a photorealistic, rendered 3D surface. Results from this work can be found here. Leading student researcher: Habib Fathi.

 

GRS: PROGRESSIVE SITE MODELING WITH VIDEOGRAMMETRY August 2009 – July 2011

NSF Grant #0943112, PI: Ioannis Brilakis, $36,978

This project was a supplement to the project above and aimed to engage an underrepresented graduate student, Ms. Stephanie German, in the core research of the parent grant. Her scope of work was including the activities needed to validate the point pair tracking and relative geo-referencing methods proposed originally. The supplement was requested due to an additional predecessor research activity (automate camera system calibration) that was discovered to be necessary after the first 6 months of tests, and was not taken into account in the original budget. Leading student researcher: Stephanie German.

 

IREE: AUTOMATED VISION TRACKING OF PROJECT RELATED ENTITIES August 2007 – August 2011

NSF Grant #0738417, PI: Ioannis Brilakis, $37,250

This international collaboration sent 3 US students to the Aristotle University of Thessaloniki (AUTh), Greece, for 4 months. The students tested the tracking method invented and prototyped in the project above on several types of sites of the Egnatia Odos motorway project, such as cantilevered bridge construction, tunnel face excavation, and interchange construction. These sites were under heavy equipment, personnel and materials traffic. Egnatia Odos is an $8 billion, 670km project that aims to create a central East-West artery to connect Turkey in the east with the Ionian Sea port of Igoumenitsa in the west. The foreign collaborator of this project, Pr. Demos Angelides, is the Chairman of the Civil and Environmental Engineering Department at AUTh and acted as the host and local advisor for our students. Results from the work can be foundhere. Leading student researcher: Gauri Jog.

 

AUTOMATED VISION TRACKING OF PROJECT RELATED ENTITIES August 2006 – August 2011

NSF Grant #0625643, PI: Ioannis Brilakis, $299,739

This research aimed to design an automated vision tracking method that reports the 4D location (spatial coordinates and time) of distinctly shaped, project related entities, such as construction equipment, personnel, and materials of standard sizes and shapes. Under this method, two or more self-calibrated, outdoor wireless video cameras are initially placed at a project site and collect video-streams. Using construction materials and shapes visual recognition techniques, each project related entity on the cameras' field of view is identified as an "interesting" pattern to track. Established tracking tools are then used in each subsequent frame of the video stream to track the movement of the identified "interesting" entity while it operates within the cameras' viewing spectrum. Results from this work can be found here. Leading student researcher: Man Woo Park.

 

REMOTE WIRELESS COMMUNICATIONS FOR CONSTRUCTION MANAGEMENT

The objective of this research was to design of a long-distance, wireless communications model suitable for data exchange between construction sites and engineering headquarters. Under this model, common types of electronic construction data can be exchanged in a fast and efficient manner, and construction site personnel can interact and share knowledge, information and electronic resources with the office staff. Results from this work can be found here.

 

MULTI MODAL RETRIEVAL OF CONSTRUCTION SITE IMAGES FROM MODEL BASED SYSTEMS

The objective of this research was to develop an all inclusive model for construction image indexing and retrieval. This model is based on Content Based Image and Video Retrieval principles while taking into account the specific characteristics and needs of the Construction Industry. Results from this work can be found here.

 

VISION BASED RETRIEVAL OF CONTEXTUAL PROJECT INFORMATION FOR RAPID ON-SITE DECISION-MAKING IN CONSTRUCTION, INSPECTION, AND MAINTENANCE

The objective of this research was to investigate the requirements of an automated vision based technique to retrieve contextual project information for supporting rapid on-site decision-making in construction, inspection, and maintenance tasks. This can be achieved by designing and implementing a pattern recognition model that allows the identification of construction entities and materials visible in a user's field of view at a given time. Results from this work can be found here.