(A)I-COMPLI: AI Supported Automated Construction compliance inspection
The issue of non-compliance with regulations, defects and poor quality during on- site construction is an issue internationally. Forinstance, defects in up to 100,000 apartment buildings in Ireland are set to cost the state at least €2.5 Billion. Current on-siteinspection regimes are largely visual, manual and analogue with error rates of up to 30% reported. They are also undertaken on a sampling basis with large portions of building elements not formally inspected, which in turn has been linked to operative behavioralissues contributing to poor quality.
This project aims to seed the disruption of visual construction inspections through the application of Artificial Intelligencetechniques based on Computer Vision and Machine Learning via camera technology. Utilizing reality capture (from real and lab-based construction examples) our system will be trained to identify and prompt potential issues of non-compliance for furtherevaluation. The solution will provide a proof of concept for the enhancing of visual inspections through a time and location specific digital solution.
When developed the solution can reduce defects, associated costs and impacts by automatically flagging potential issues andtransforming compliance behavior. In turn it can disrupt the sector and support digitization while delivering positive social, economic and environmental impacts.
Collaborators:
Evercam Construction Cameras
Collaborators
Principal Investigator

Dr. Mark Mulville
Technological University Dublin
Head of the School of Surveying and Construction Innovation in TU Dublin.
A Building Surveyor and Architectural Technologist, Mark’s research interests include the delivery of long term building performance in the context of a changing climate, energy efficiency, comfort, health and well-being in buildings and the optimisation of construction processes utilising digital technologies.
Mark is currently principal investigator on the DTIF funded A-EYE project developing a construction visualisation platform to support construction productivity. Mark is also principal investigator on the SEAI funded CC-DORM project examining overheating risk in dwellings in the context of large scale retrofit and climate change. Mark previously led the Building Performance research cluster in the University of Greenwich (UK).