Surrogate modelling as enabling methodology for predictive Digital Twins in geohazards

Summary In hole enlargement while drilling (HEWD) operations, underreamers are used extensively to enlarge the pilot hole. Reamer wipeout failure can cause additional bottomhole assembly (BHA) trips, which can cost operators millions of dollars. Excessive reamer shock and vibration are leading causes of reamer wipeout; therefore, careful monitoring of reamer vibration is important in mitigating such a risk. Currently, downhole vibration sensors and drilling dynamics simulations (DDSs) are used to comprehend and reduce downhole vibration, but vibration sensors cannot be placed exactly at the reamer to monitor the vibrations in real time. DDSs are difficult to calibrate and are computationally expensive for use in real time; therefore, the real-time reamer vibration status is typically unknown during drilling operations. A process digital twin using a hybrid modeling approach is proposed and tested to address the vibration issue. Large amounts of field data are used in advanced DDSs to calibrate the HEWD runs. For each HEWD section, calibrated DDSs are performed to comprehend the downhole vibration at the reamer and downhole vibration sensors. A surrogate regression model between reamer vibration and sensor vibration is built using machine learning. This surrogate model is implemented in a drilling monitoring software platform as a process digital twin. During drilling, the surrogate model uses downhole measurement while drilling (MWD) data as inputs to predict reamer vibration. Wipeout risk levels are calculated and sent to the operators for real-time decision-making to reduce the possibility of reamer wipeout. Large volumes of reamer field data, including field recorded vibration and reamer dull conditions were used to validate the digital twin workflow. Then, the process digital twin was implemented and tested in two reamer runs in the Gulf of Mexico. A downhole high-frequency sensor was placed 8 ft above the reamer cutting structure in one field run, and the recorded sensor vibration data and corresponding reamer dull conditions showed a very good match with the real-time digital twin predictions in a low-vibration scenario. Cases in high vibration are needed to fully validate the feasibility and accuracy of the digital twin. State-of-the-art downhole sensors, DDS packages, large amounts of field data, and a hybrid approach are the solutions to building, calibrating, and field testing the reamer digital twin to ensure its effectiveness and accuracy. Such a hybrid modeling approach can not only be applied to reamers but also to other critical BHA components.

We develop novel approaches to enhance the functionality and efficiency of digital twins through deep learning techniques. Digital twins, sophisticated virtual models of physical systems, serve as dynamic counterparts, mirroring real-world entities' behavior and performance. Their primary role includes real-time monitoring, flaw detection, automated control, and proactive maintenance. To perform such tasks it is necessary to use various methods from scientific computing to combine data with physics models. Furthermore, to assess the reliability of a digital twin, it is important to not only make predictions, but also quantify the uncertainty of estimations and predictions. In this regard, Bayesian methods serve as a natural framework to pose the problems. However, the complexity and real-time operation requirements pose significant computational challenges. The thesis explores the integration of deep learning in scientific computing for enabling digital twins. The focus is on overcoming the computational hurdles in real-time data assimilation and inverse problem-solving through surrogate modeling. Deep learning, with its capability to handle high-dimensional functions in both deterministic and stochastic settings is shown to be a viable solution to these challenges. This work investigates several key areas, beginning with a brief introduction to digital twins, data assimilation, and inverse problems. It proceeds to elaborate on the role of Bayesian inversion problems in static and dynamic contexts, highlighting the importance of uncertainty quantification. The focus is on the challenges of real-time computation and the potential of Bayesian methods, despite their computational intensity. The thesis investigates deep learning architectures like dense and residual neural networks, convolutional neural networks, recurrent neural networks and transformers. The core contributions are presented in the four main chapters of the thesis: 1) Reduced Order Modeling for Parameterized Time-Dependent Partial Differential Equations: Introducing a deep learning framework for solving parameterized PDEs using spatially and memory-aware neural networks, demonstrated on linear advection and incompressible Navier-Stokes equations. 2) Markov Chain Generative Adversarial Neural Networks: Enhancing Bayesian inversion for static problems with sparse observations using a unique algorithm combining MCMC and GANs, demonstrated on a Darcy flow problem and pipe flow leakage detection. 3) Probabilistic Digital Twin for Leak Localization: Developing a probabilistic framework using supervised Wasserstein Autoencoders for leak localization in water distribution networks, tested on several network models. 4) The Deep Latent Space Particle Filter: Proposing a real-time nonlinear data assimilation method for dynamic PDEs with uncertainty quantification, using transformer-based dimensionality reduction and time-stepping in particle filters, demonstrated on various dynamic systems. This thesis contributes significantly to digital twins, addressing computational limitations and opening new avenues for practical applications in various industries.

A digital twin (DT), with the components of a physics-based model, a data-driven model, and a machine learning (ML) enabled efficient surrogate model, behaves as a virtual twin of the real-world physical process. In terms of Laser Powder Bed Fusion (L-PBF) based additive manufacturing (AM), a DT can predict the current and future states of the melt pool and the resulting defects corresponding to the input laser parameters, evolve itself by assimilating in-situ sensor data, and optimize the laser parameters to mitigate defect formation. In this paper, we present a deep neural operator enabled DT framework for closed-loop feedback control of the L-PBF process. This is accomplished by building a physics-based computational model to accurately represent the melt pool states, an efficient Fourier neural operator (FNO) based surrogate model to approximate the melt pool solution field, followed by a physics-based procedure to extract information from the computed melt pool simulation that can further be correlated to the defect quantities of interest (e.g., surface roughness). An optimization algorithm is then exercised to control laser input and minimize defects. On the other hand, the constructed DT also evolves with the physical twin via offline finetuning and online material calibration. For instance, the probabilistic distribution of laser absorptivity can be updated to match the real-time captured thermal image data. Finally, a probabilistic framework is adopted for uncertainty quantification. The developed DT is envisioned to guide the AM process and facilitate high-quality manufacturing in L-PBF-based metal AM.

Data-driven uncertainty quantification in computational human head models

Machine learning based digital twin for dynamical systems with multiple time-scales

Computational simulation approaches have long been a cornerstone of design in marine and offshore structures. More recently, they have also become essential for optimizing operations and ensuring sustainability. The digital twin (DT) framework has enabled to bridge the gap between simulation and its practical application in controlling, monitoring, and operating marine assets. However, the applications where complex non-linear problems are solved in real or close to real-time are still limited. It is proposed that this could be overcome by the recent rise of artificial intelligence (AI) and machine learning methods, which are promising technologies that can make current computational models more versatile. Namely, data-driven surrogate models or neural networks embedded in numerical algorithms enable the relegation of computationally complex calculations into simple, functional forms, opening a pathway to a new generation of digital twins. Memory-augmented neural networks offer a viable solution for applications involving memory-dependent problems, such as ship motion control or material fracture. These neural network models operate incrementally, much the same as the computational models they are inserted to. This work explores a critical yet underexamined aspect of these incremental machine learning models that is important when considering the pathway to real-time digital twins for ship motion control or non-linear structural analysis.

Biopharma Is Going Digital … Bit by Bit

The move towards advanced manufacturing and Industry 4.0 is fed by increased demand for speeding up innovation, increasing flexibility, improving maintenance, and becoming more customized while saving on the total cost of operations. This is accompanied by increased dependence on virtual product and process development, data‐driven processes, and product knowledge. Other characteristics, affecting modern design, include big data intelligence in product, process, and maintenance. New technologies that empower data analytics include added manufacturing, flexible manufacturing, robotics, sensor technology, smart value/supply chains, and industrial information backbones. This paper is about surrogate models, also called digital twins, that provide an important complementary capacity to physical assets. Digital twins capture past, present, and predicted behavior of physical assets. Digital twin models are updated periodically to represent the current state of physical assets. This distinguishes digital twins from conventional simulations in that sensor data can continuously feed them. The type of curated information on the state of physical asset's history depends on how digital twins are used. For example, if a digital twin is used for fault classification, the history captured is operational data from equipment in healthy and faulty states. We provide here a review of digital twins, with an emphasis on Industry 4.0 applications.

Propelled by recent advances in Mobile Edge Computing (MEC) and the Internet of Things (IoT), the digital twin technique has been envisioned as a de-facto driving force to bridge the virtual and physical worlds through creating digital portrayals of physical objects. In virtue of the flourishing of edge intelligence and abundant IoT data, data-driven modelling facilitates the implementation and maintenance of digital twins, where simulations of physical objects are usually performed based on Deep Neural Networks (DNNs). A significant advantage of adopting digital twins is to enable decisive prediction on the behaviours of objects in near future without waiting for that really happen. To provide accurate predictions, it is vital to keep each digital twin synchronized with its physical object in real-time. However, it is challenging to maintain the real-time synchronization between a digital twin and its physical object due to the dynamics of physical objects and sensing data drift over time, i.e., the live data from a physical object diverge from the model training data of its digital twin. To address this critical issue, continual learning is a promising solution to retrain models of digital twins incrementally. In this paper, we investigate digital twin synchronization issues via continual learning in an MEC environment, with the aim to maximize the total utility gain, i.e., the enhanced model accuracy. We study two novel optimization problems: the static digital twin synchronization problem per time slot and the dynamic digital twin synchronization problem for a finite time horizon. We first formulate an Integer Linear Program (ILP) solution for the static digital twin synchronization problem when the problem size is small; otherwise, we develop a randomized approximation algorithm at the expense of bounded resource violations for it. We also devise a deterministic approximation algorithm with guaranteed performance for a special case of the static digital twin synchronization problem. We thirdly consider the dynamic digital twin synchronization problem by proposing an efficient online algorithm for it. Finally, we evaluate the performance of the proposed algorithms for continuous digital twin synchronization through simulations. Simulation results show that the proposed algorithms are promising, outperforming counterpart benchmarks by no less than 13.2%, in terms of the total utility gain.

Credibility consideration for digital twins in manufacturing

This paper elaborates on the large number of Uncertainty Quantification (UQ) techniques that have been proposed to enhance the reliability and the fidelity of Digital Twins that are increasingly finding applications in domains like Robotics and Smart Manufacturing. Digital twins are virtual duplicates or virtual models of a physical asset; they use advanced techniques such as data analytics and simulation-driven methods. However, the development and use of these advanced systems are plagued by a host of uncertainties, which are mainly introduced from sensor noise, intermittent connectivity, biases from data processing, and model abstractions and simulation stochasticity. Such uncertainties can be quantified by methods such as frequentist statistics, interval analysis, Bayesian inference, and random sampling. The mapping is important in gaining insights into these UQ methods and their associated advantages and limitations and the mitigation guidelines are to be used throughout the Digital Twin pipeline. UQ at its core involves real-time adaptive control in dynamically changing environments that leverage state awareness towards responsive action within predictive control models and feedback systems. In addition, machine learning algorithms support the ability to make better decisions from the identification of patterns in historical data to make plans for responsive trajectories of robots. UQ further allows the collaboration of human and machine, giving early warnings on anomalies and risks that enhance visibility which further fosters coordination and communication during disruptive situations. Robust development of digital twins for robotics and manufacturing relies on integrated UQ practices. The current review provides best practices, insights, and guidelines on the application of UQ across modeling, control strategies, and collaborative workflows aimed at delivering actionable and reliable insights from digital twin simulations, analytics, and decision support.

Since each cancer has its own unique characteristics, each one can respond differently to the same treatments. Therefore, the creation of a digital twin (DT) of cancer can assist us in predicting the evolution of an individual's cancer through modeling each tumor's characteristics and response to treatment. Hence, we propose to take advantage of new advances in computational approaches and combine mechanistic, machine learning, and stochastic modeling approaches to create “My Virtual Cancer", a DT platform. To establish a personalized DT, we use patient-specific data for parameter estimations, sensitivity analysis, and uncertainty quantification. For each patient, we will estimate the values of parameters of their QSP model using the patient's data. We perform a multi-dimensional sensitivity analysis and uncertainty quantification on the mechanistic model to find a set of critical interactions and predict the intervals of confidence. Since this QSP model includes the data-driven mechanistic model of cells and molecules' interaction networks, one of the ultimate results of this DT would be the prediction of evolution of tumors.

A digital twin is a powerful new concept in computational modelling that aims to produces a one-to-one mapping of a physical structure, operating in a specific context, into the digital domain. This technology could therefore provide improved and robust decision making for asset management. Although the applications of digital twins vary, this paper focuses on digital twins for structural dynamic systems. A key consideration in developing a digital twin is in the construction of a workflow that defines decisions and interactions within the modelling framework. This process will generally be bespoke to specific applications, however key principles will apply. Furthermore, a workflow will provide a methodology for identifying poor predictive performance and systematically improving predictions via optimal decision making. In this paper a three storey building structure is introduced as a case study in order to motivate the challenges and technologies required of a digital twin. The context of this case study is to develop a digital twin of the building structure that consistently predicts the acceleration response of the three floors given an unknown structural state, caused by a contact nonlinearity between two floors. This reflects realistic challenges for a digital twin in that the physical twin will degrade with age, and its response may change under various loading scenarios, unforeseen in the initial model development phase. Key elements within a potential workflow for this application are discussed. These include indicating when model updating schemes become problematic and how augmenting physics-based models with a data-based component can provide information about poor predictive performance. These techniques are linked to hybrid testing, as a potential method for improving model development based on the physical structure in a controlled offline manner. Finally, the impact of these procedures are discussed for model based control methods in terms of vibration attenuation performance, but also robustness against model uncertainties and external disturbances. The workflow and key technologies investigated in this specific case study are expected to outline the general processes that apply to digital twins more broadly, and provide a clearer understanding of how a digital twin should be implemented.KeywordsDigital twinValidationHybrid testingRobust control

Global sensitivity analysis in high dimensions with PLS-PCE

Abstract. Technology has advanced and progressed tremendously, and the term city is being elevated to a new level where the smart city has been introduced globally. Recent developments in the concept of smart city have led to a renewed interest in Digital Twin. Using precise Building Information Modelling (BIM) consolidated with big data and sensors, several attempts have been made to establish digital twin smart cities. In recent years, several researchers have sought to determine the capability of smart city and digital twin for various taxonomies such as development and urban planning purposes, built environment, manufacturing, environmental, disaster management, and healthcare. Despite being beneficial in many disciplines, especially in manufacturing, built environment, and urban planning, these existing studies have shown a lack of aspect in terms of emergency or disaster-related as opposed to the elements mentioned above. This is because the researcher has not treated emergencies or disasters in much detail. Therefore, an extensive review on smart city, digital twin, BIM and disaster management and technology that revolves around these terms were summarised. In general, 39 articles from prominent multidisciplinary databases were retrieved over the last two decades based on the suggested PRISMA workflow. These final articles were analysed and categorised into four themes based on the research content, gist, and keywords. Based on the review of 39 articles related to smart city, digital twin and BIM, a workflow for the smart city digital twin and the conceptual framework for indoor disaster management was proposed accordingly. The establishment of smart city digital twins solely for an indoor emergency can be beneficial to urbanites, and it could provide numerous benefits for enhanced situation assessment, decision making, coordination, and resource allocation.