Back

Artificial Intelligence in Construction: Predictive Models and Concrete Pour Detection

Rodrigo Alzola
2025-03-11
5 Minutes read
Artificial Intelligence in Construction: Predictive Models and Concrete Pour Detection

Abstract

The construction industry is undergoing a digital transformation, leveraging artificial intelligence to enhance efficiency and decision-making. ObraLink integrates AI to revolutionize construction processes, focusing on predictive models for concrete consumption and automated detection of concrete pouring through computer vision. By analyzing historical project data, ObraLink provides accurate forecasts that optimize planning and reduce estimation errors. Additionally, its AI-powered monitoring system detects construction progress in real-time, linking visual data to BIM workflows for precise tracking and analytics. This innovative approach empowers construction professionals to streamline operations, minimize waste, and embrace a smarter, data-driven future

The goal of ObraLink is to help people and companies incorporate technological processes into their workflow, improving decision-making and empowering them to build the future human environment. One of the tools we use to help digitalize the construction world is artificial intelligence.

How?

Image that shows a construction team reviewing plans and schedules on a work table, illustrating the importance of planning in construction

Image that shows a construction team reviewing plans and schedules on a work table, illustrating the importance of planning in construction

One of our artificial intelligence applications is the generation of predictive models for concrete pouring. Thanks to the collection of historical data from construction projects of different sizes over the past 4 years, we have developed forecasting models to predict how much concrete will be used weekly. These models allow us to generate suggestions for the planner, and thus reduce the error in the estimation of concrete pouring. This helps those who plan to reduce the difference between what is planned and what is actually executed.
Another way in which Obralink integrates artificial intelligence in construction is through the detection of concrete pouring using computer vision. Through images captured by the camera of our Cibot, we are able to identify where and when there has been a progress in the concreting of a construction site. This way we can efficiently detect concrete pouring, and also, thanks to the three-dimensional coordinates of the building, we can link the advances of the images with their location in the construction plan. By integrating this technology with the BIM flows of the platform, we are able to, from the detection of the advance, cubiculate the progress and apply analytics to the construction site, in order to monitor the project optimally. It is important to comment that, given the complexities inherent to a construction site (e.g., visual pollution), it is still necessary to perform manual review of what is detected (human-in-the-loop), seeking to improve the labeling process for future models, that will be integrated with our software.
a) Image of the Cibot camera, b) Automatic detection of concrete in the plan

a) Image of the Cibot camera, b) Automatic detection of concrete in the plan

The best thing is that we are just beginning, there are still infinite ways in which technology and artificial intelligence can benefit the construction industry! And these first examples show that technological advances and artificial intelligence are perfectly applicable in the construction industry. That's why at Obralink we are excited to think about how the future of construction will be and see how far our developments can go.
If you are interested in learning more about the application of artificial intelligence in construction, contact us here

About the author

Go to author profile
Rodrigo Alzola

Rodrigo Alzola

Rodrigo Alzola is the head of Artificial Intelligence at Obralink, with more than 6 years of experience in computer vision and machine learning.