In the ever-evolving landscape of the Architecture, Engineering, Construction, Operations, and Management (AECOM) industries, a groundbreaking approach is emerging to tackle long-standing challenges, particularly in geotechnical design and construction. Researchers, led by Dafydd Cotoarbă from the Georg Nemetschek Institute Artificial Intelligence for the Built World at the Technical University of Munich, have introduced a novel framework that promises to revolutionize decision-making processes in the face of uncertainty.
The digital twin (DT) approach has been gaining traction as a solution to the complexities of AECOM projects. However, traditional DTs have a significant limitation: they rely on deterministic models that require precise input parameters. This rigidity can lead to inaccuracies, as it fails to account for the inherent uncertainties in geotechnical projects. “The uncertainties in geotechnical design and construction are substantial and multifaceted,” explains Cotoarbă. “Our proposed probabilistic digital twin (PDT) framework addresses this by incorporating and propagating these uncertainties throughout the modeling process.”
The PDT framework is designed to integrate various sources of uncertainty, including aleatoric, data, model, and prediction uncertainties. This comprehensive approach ensures that site-specific conditions are accurately reflected as new information becomes available. The framework leverages Bayesian methods for model updating, making it a dynamic and adaptive tool for real-world engineering workflows.
One of the most compelling aspects of this research is its potential commercial impact, particularly in the energy sector. Geotechnical design and construction are critical components of energy infrastructure projects, such as wind farms, oil and gas installations, and renewable energy plants. The ability to make informed decisions under uncertainty can lead to more efficient project execution, reduced costs, and improved safety outcomes.
The effectiveness of the PDT framework was demonstrated through an application to a highway foundation construction project. The results showcased the framework’s potential to integrate existing probabilistic methods, enhancing decision-making and project outcomes. “By embedding these methods within the PDT framework, we lower the barriers to practical implementation,” says Cotoarbă. “This makes probabilistic approaches more accessible and applicable in real-world engineering workflows.”
The research, published in the journal ‘Data-Centric Engineering’ (translated to English as ‘Data-Centric Engineering’), represents a significant step forward in the field of geotechnical design and construction. As the energy sector continues to evolve, the need for robust and adaptable tools becomes ever more critical. The PDT framework offers a promising solution, paving the way for more resilient and efficient infrastructure projects.
In the words of Cotoarbă, “This is not just about improving models; it’s about transforming how we approach engineering challenges. It’s about making our infrastructure more reliable, our decisions more informed, and our projects more successful.” As the industry continues to grapple with uncertainty, the PDT framework stands as a beacon of innovation, guiding the way towards a more secure and sustainable future.