ŷ���ڱ���Ƶ

Naiara Rodrigues Tonin, a PhD student in the Structural Engineering and Structural Mechanics program, conducts tests related to this project.

Researchers at ŷ���ڱ���Ƶ Boulder are developing an app that could reliably and quickly predict whether batches of concrete made at construction sites are safe. If successful, the work could usher in a new era of building that is faster, more cost effective and safer overall for everyone.

The work is still in its early stages and is funded by a seed grant from the Engineering Education and AI-Augmented Learning Interdisciplinary Research Theme within the College of Engineering and Applied Science. 

Assistant Professor Mija Hubler of civil, environmental and architectural engineering said the goal of the project was to develop an app that could collect and analyze sample images of concrete for possible defects using machine learning techniques based on composition, fault lines and visual clues.

“To do that today, we have to send samples to the lab where it is then destroyed to be analyzed – so it isn’t a very efficient process in many ways,” she said. “We are hoping to develop something where you could cut a sample open on site, take a picture and understand how that batch will perform mechanically.”

Hubler said the approach is similar to medical techniques, in which  doctors examine images of bones or organs to make assessments with increasingly sophisticated tools and techniques. 

However, concrete is a much less homogeneous material, which makes assessment tricky. And users of such an app would need at least some basic education about machine learning to understand the inherent uncertainty in the predictions and how to proceed with them.

“We are talking about – essentially – a smart tool here. These kinds of tools and skills are going to become more common in construction over the next 20 years, especially with the introduction of autonomous vehicles,” she said. “We quickly realized then that it is more of a question of education for everyone on the site. How do we teach these skills? How much does someone need to know about machine learning to use these tools? That is why it fits well with the Engineering Education and AI-Augmented Learning Interdisciplinary Research Theme.”

To address those kinds of questions, Hubler is working with computer science Assistant Teaching Professor Geena Kim. Kim said that while artificial intelligence and machine learning are increasingly common in her field, many other branches of engineering are just starting to use those concepts. She added that the scale and size of data sets in civil engineering make for interesting challenges when creating the needed algorithms, testing with students and understanding the results for broad applications.

“We need to get more data and observations to really understand how people will interact with this app and what their personal experience with AI and machine learning needs to be to use it properly,” she said. “This work will also help with our understanding of these concepts in curriculum and workforce development over time.”

Hubler said the team will continue to refine their approach, while also seeking collaborators at ŷ���ڱ���Ƶ Boulder and beyond.

“The primary way we assess and track our infrastructure in America is through visual inspections, so this kind of tool would be quite powerful,” she said.