#GESmart: To transform underground development, current practices should evolve to the next level: the digital twin’

The GE Smart Geotechnics conference is taking place next week on Thursday 3 October!

Don't miss out on?this?essential gathering for anyone involved in geotechnical monitoring, data management and visualisation.

The conference will provide?up-to-date insights into projects, technologies and instrumentation, data systems and modelling software, as well as the use of the cloud, machine learning and AI.

Don't wait any longer – secure your place today!

***Ground Engineering subscribers are entitled to a ￡100 discount***

At the GE Smart Geotechnics conference, University of Birmingham’s senior lecturer in digital engineering Jelena Ninic will discuss the automated reconstruction of digital twins for underground infrastructure.

Society is increasingly recognising the need to engineer future cities to be resilient to natural and human-made hazards, while promoting well-being, sustainable development, resource security and inclusive growth.

As there is limited land in many parts of the world, there is an opportunity to utilise the often-overlooked underground space for the development of liveable cities.

Achieving global decarbonisation goals and resilient underground infrastructure requires proactive processes, and the development of reliable digital technologies makes this shift possible.

The adoption of building information modelling (BIM) across the construction industry has improved delivery and performance, fostered collaboration and innovation, and introduced opportunities for a new level of automation.

Although BIM is now widely used, the current adoption level still does not properly exploit the digital revolution. BIM is advanced but siloed, as users believe that their competitive advantage is protected by keeping their models closed. However, in practice, this hinders data sharing and knowledge aggregation that would benefit a range of stakeholders

To transform sustainable underground development, current practices should evolve to the next level: the digital twin - a real-time single source of truth, to bring together data from disparate systems, optimise design and enhance productivity.

But why are digital twins for underground not out there yet?

There are many challenges and knowledge gaps to be filled: the lack of generalised, scalable approaches for digital modelling of the whole life cycle (BIM for underground is today generally applied on a case-study basis for the design phase); lack of standardisation and interoperability; integration of real-time monitoring data; integration with advanced assessment tools and holistic assessment approaches; and underexploited cross-sectoral collaboration in the development of methods and applications.

In the complex and uncertain environment of underground systems, managing complexity is crucial for effective decision-making, requiring reliable yet user-friendly tools. Digital twins can be complex and difficult to construct, which hinders their wider adoption.

To address this issue, our research is concerned with automation in the reconstruction and execution of digital twins. To achieve a high level of automation, we propose solutions for the automated reconstruction of digital models, integrated assessment, real-time design optimisation in BIM, virtual control of construction, and the application of machine learning (ML) for inspection and maintenance.

The first step is to reconstruct and semantically enrich the digital model. We can reconstruct ground and underground structural components from diverse data sources using parametric object modelling techniques and soft computing methods for semantic enrichment.

Our solution incorporates automation in the reconstruction of digital twins based on BIM technology, including automated component detection, model reconstruction in BIM, condition detection, localisation and quantification, and mapping conditions onto the digital models. This allows us to not only create but also easily update information-rich digital models that can then be used for assessment, analysis, scenario exploration and more.

The second step is the analysis of the design. We demonstrate how the seamless integration of BIM and high-fidelity numerical models can transform geotechnical project planning through efficient, computationally enhanced decision-making, maximising flexibility, ensuring the robustness and reducing both costs and CO2 emissions. This efficient integration of analysis in the early design phase allows for more holistic, multi-objective optimisation of solutions and the implementation of changes when the cost of those changes is minimal. By adopting this approach, we can achieve more sustainable, cost-effective, and adaptable design solutions for geotechnical projects.

Moving to a higher level of automation, ML-based metamodels integrated within digital twins can be employed for real-time design assessment and process control during construction. This is achieved by training metamodels during the offline phase, and then, during construction, updating and fusing them with monitoring data for more reliable predictions. Since metamodels can execute in real time, they can optimise construction and operational parameters on the fly, minimising the environmental impact of construction.

Finally, during the operation phase, the digital twin can be continuously updated with monitoring and condition data using computer vision and deep learning methods in an automated way. This enables the creation of a living digital replica of the asset, which can be used for effective and proactive asset maintenance.

To discover more about the 2024 programme click HERE .