Research interests

  • Digital Twins of the Built Environment
  • Building Information Modeling
  • Construction Automation
  • Damage and Defects Assessment
  • Computer Vision
  • Machine Learning
  • Augmented Reality
  • Product Lifecycle Management (PLM)

The overarching goals of the Digital Infrastructure Systems lab are to digitize infrastructure under three performance metrics: a) cost reduction of labor intensive engineering tasks, (b) energy-efficiency and resilience and (c) student learning outcomes of Digital Construction using Artificial Intelligence. Digital twinning applications span across industries, i.e. aerospace, manufacturing, energy, construction and engineering.

Digital Twins of the Built Environment

Machine learning and optimization algorithms have been rapidly developed and improved during the last decades with deep learning being one of the most prominent fields. Could you imagine the real world impact of applying these algorithms on data from physical infrastructure assets? This would fundamentally change the construction industry and would allow generating designs superior to human design, developing thorough maintenance and energy efficiency strategies by generating their up-to-date copy, the Digital Twin. The common goal of our research is to automatically generate those very accurate digital models of existing infrastructure which can then be used in a variety of tasks to obtain much more accurate results compared to simplified designs.

Selected Publications:

Automated Visual Inspections

There are more than 600,000 bridges across the United States, with 8% of those being structurally deficient. Current visual bridge inspections are manual with many bridge inspection records missing. Machine learning methods will be deployed to improve the current understanding of bridge conditions and predict their life.

Infrastructure Computer Vision

The lab specializes on spatial understanding applications using deep learning algorithms to the built environment, including post-earthquake damage assessment and generative design.

Selected Publications:

Research Sponsors

Scroll to Top