Indranil Pan's research while affiliated with Imperial College London and other places

Publications (27)

Preprint
Full-text available
Understanding where normal faults are is critical to an accurate assessment of seismic hazard, the successful exploration for and production of natural (including low-carbon) resources, and for the safe subsurface storage of CO2. Our current knowledge of normal fault systems is largely derived from seismic reflection data imaging intra-continental...
Article
Full-text available
We introduce a novel rule-based approach for handling regression problems. The new methodology carries elements from two frameworks: (i) it provides information about the uncertainty of the parameters of interest using Bayesian inference, and (ii) it allows the incorporation of expert knowledge through rule-based systems. The blending of those two...
Article
Full-text available
Reduced-order models (ROMs) are computationally inexpensive simplifications of high-fidelity complex ones. Such models can be found in computational fluid dynamics where they can be used to predict the characteristics of multiphase flows. In previous work, we presented a ROM analysis framework that coupled compression techniques, such as autoencode...
Preprint
Full-text available
Understanding where normal faults are is critical to an accurate assessment of seismic hazard, the successful exploration for and production of natural (including low-carbon) resources, and for the safe subsurface storage of CO2. Our current knowledge of normal fault systems is largely derived from seismic reflection data imaging intra-continental...
Chapter
We highlight the key pillars of urban energy systems which would leverage on AI and digital technologies for a low carbon future. We summarise a couple of real world applications where optimisation, intelligent control systems and cloud-based infrastructure have played a transformative role in improving system performance, cost-effectiveness and de...
Preprint
Full-text available
In our previous work, we introduced the rule-based Bayesian Regression, a methodology that leverages two concepts: (i) Bayesian inference, for the general framework and uncertainty quantification and (ii) rule-based systems for the incorporation of expert knowledge and intuition. The resulting method creates a penalty equivalent to a common Bayesia...
Preprint
As we transition from fossil fuel to renewable energy, negative emission technologies, such ascarbon capture and storage (CCS), can help us reduce CO2 emissions. Effective CO2 storage requires: (1) detailed site characterization, (2) regular, integrated risk assessment, and (3) flexible design and operation. We believe that recent advances in machi...
Preprint
Full-text available
Reduced-order models (ROMs) are computationally inexpensive simplifications of high-fidelity complex ones. Such models can be found in computational fluid dynamics where they can be used to predict the characteristics of multiphase flows. In previous work, we presented a ROM analysis framework that coupled compression techniques, such as autoencode...
Preprint
Full-text available
Recent advances in machine learning, coupled with low-cost computation, availability of cheap streaming sensors, data storage and cloud technologies, has led to widespread multi-disciplinary research activity with significant interest and investment from commercial stakeholders. Mechanistic models, based on physical equations, and purely data-drive...
Article
Recent advances in machine learning, coupled with low-cost computation, availability of cheap streaming sensors, data storage and cloud technologies, has led to widespread multi-disciplinary research activity with significant interest and investment from commercial stakeholders. Mechanistic models, based on physical equations, and purely data-drive...
Article
Understanding the internal structure of our planet is a fundamental goal of the earth sciences. As direct observations are restricted to surface outcrops and borehole cores, we rely on geophysical data to study the earth's interior. In particular, seismic reflection data showing acoustic images of the subsurface provide us with critical insights in...
Article
Non-intrusive reduced-order models (ROMs) have recently generated considerable interest for constructing computationally efficient counterparts of nonlinear dynamical systems emerging from various domain sciences. They provide a low-dimensional emulation framework for systems that may be intrinsically high-dimensional. This is accomplished by utili...
Article
Full-text available
Multicomponent polymer systems are of interest in organic photovoltaic and drug delivery applications, among others where diverse morphologies influence performance. An improved understanding of morphology classification, driven by composition-informed prediction tools, will aid polymer engineering practice. We use a modified Cahn–Hilliard model to...
Preprint
Understanding the internal structure of our planet is a fundamental goal of the Earth Sciences. As direct observations are restricted to surface outcrops and borehole cores, we rely on geophysical data to study the Earth’s interior. Especially, seismic reflection data showing acoustic images of the subsurface, provide us with critical insights into...
Article
Full-text available
In chemical process engineering, surrogate models of complex systems are often necessary for tasks of domain exploration, sensitivity analysis of the design parameters, and optimization. A suite of computational fluid dynamics (CFD) simulations geared toward chemical process equipment modeling has been developed and validated with experimental resu...
Preprint
Seismic facies analyses are fundamental to the study of sedimentary, tectonic and magmatic systems using seismic reflection data. These analyses generally assume that seismic facies are: (1) well defined, (2) distinct and (3) prevalent patterns in the data. Here, we examine these assumptions critically. First, we demonstrate how to extract the main...
Preprint
We introduce a novel rule-based approach for handling regression problems. The new methodology carries elements from two frameworks: (i) it provides information about the uncertainty of the parameters of interest using Bayesian inference, and (ii) it allows the incorporation of expert knowledge through rule-based systems. The blending of those two...
Preprint
Full-text available
As Internet of Things (IoT) technologies enable greater communication between energy assets in smart cities, the operational coordination of various energy networks in a city or district becomes more viable. Suitable tools are needed that can harness advanced control and machine learning techniques to achieve environmental, economic and resilience...
Preprint
Non-intrusive reduced-order models (ROMs) have recently generated considerable interest for constructing computationally efficient counterparts of nonlinear dynamical systems emerging from various domain sciences. They provide a low-dimensional emulation framework for systems that may be intrinsically high-dimensional. This is accomplished by utili...
Preprint
Multi-component polymer systems are of interest in organic photovoltaic and drug delivery applications, among others where diverse morphologies influence performance. An improved understanding of morphology classification, driven by composition-informed prediction tools, will aid polymer engineering practice. We use a modified Cahn-Hilliard model t...
Article
As Internet of Things (IoT) technologies enable greater communication between energy assets in smart cities, the operational coordination of various energy networks in a city or district becomes more viable. Suitable tools are needed that can harness advanced control and machine learning techniques to achieve environmental, economic and resilience...
Preprint
Full-text available
A suite of computational fluid dynamics (CFD) simulations geared towards chemical process equipment modelling has been developed and validated with experimental results from the literature. Various regression based active learning strategies are explored with these CFD simulators in-the-loop under the constraints of a limited function evaluation bu...
Conference Paper
Full-text available
Understanding the Earth's internal structure is one of the key challenges of geophysics. Advanced geophysical techniques, such as 3-D seismic technology, produce increasingly large datasets of the Earth's subsurface, which require significant amounts of time, experience and expertise to analyze. On the other hand, there have recently been great adv...
Article
Within the context of the Smart City, the need for intelligent approaches to manage and coordinate the diverse range of supply and conversion technologies and demand applications has been well established. The wide-scale proliferation of sensors coupled with the implementation of embedded computational intelligence algorithms can help to tackle man...
Article
Seismic interpretations are, by definition, subjective and often require significant time and expertise from the interpreter. We are convinced that machine-learning techniques can help address these problems by performing seismic facies analyses in a rigorous, repeatable way. For this purpose, we use state-of-the-art 3D broadband seismic reflection...
Article
Full-text available
Silica diagenesis has the potential to drastically change the physical and fluid flow properties of its host strata and therefore plays a key role in the development of sedimentary basins. The specific processes involved in silica diagenesis are, however, still poorly explained by existing models. This knowledge gap is addressed by investigating th...

Citations

... Perhaps the first use of an autoencoder for dimensionality reduction within a ROM framework was applied to reconstruct flow fields in the near-wall region of channel flow based on information at the wall, 30 whilst the first use of a convolutional autoencoder came 16 years later and was applied to Burgers Equation, advecting vortices and lid-driven cavity flow. 31 In the few years since 2018, many papers have appeared, in which convolutional autoencoders have been applied to sloshing waves, colliding bodies of fluid and smoke convection; 32 flow past a cylinder; [33][34][35] the Sod shock test and transient wake of a ship; 36 air pollution in an urban environment; 37-39 parametrized time-dependent problems; 40 natural convection problems in porous media; 41 the inviscid shallow water equations; 42 supercritical flow around an airfoil; 43 cardiac electrophysiology; 44 multiphase flow examples; 45 the Kuramoto-Sivashinsky equation; 46 the parametrized 2D heat equation; 47 and a collapsing water column. 48 Of these papers, those which compare autoencoder networks with POD generally conclude that autoencoders can outperform POD, 31,33 especially when small numbers of reduced variables are used. ...
... ML technics have been highly applied in this literature review. Few articles that have not used AI explicitly still consider it as an important part of their future studies (Pan et al., 2022;Zacharaki et al., 2021). AI can help DTs reach maturity and wisdom throughout the entire product lifecycle by playing two roles : ...
... The seismic volume was zero-phase processed with SEG normal polarity; i.e., a positive reflection (white) corresponds to an acoustic impedance (density × velocity) increase with depth. More details on data acquisition and pre-processing steps are provided by Wrona et al., (2019Wrona et al., ( , 2021a. ...
... In a recent work [36], the authors utilized CAE for dimensionality reduction, temporal CAE to encode the solution manifold, and dilated temporal convolutions to model the dynamics. In [52], the performance of the CAE, variational autoencoders and POD to obtain low dimensional embedding has been examined and Gaussian processes regression to implement the mapping between the input parameters and the reduced solutions. ...
... Their neural network model was trained and tested on 3D synthetic seismic data and they concluded that more complex fault structures were required in the synthetic data to enhance the mapping quality. It was demonstrated by Wrona et al. (2020) how cropped 2D samples with manually labelled faults from large 2D seismic sections from the northern North Sea could be used in a supervised learning framework for identifying faults in 2D sections. They compared two different CNN models: (1) a simple 2D CNN model that downsamples the cropped input image and classifies it, and (2) a 2D U-Net model (based on the work of Long et al. 2015;Ronneberger et al. 2015) that labels the cropped input image pixel-wise. ...
... We now apply the analysis pipeline to polymer precipitation dynamics, of importance to engineering design problems in high-performance plastics and membrane systems. The complex dynamics and rich pattern formation were considered recently by Inguva et al. (2020Inguva et al. ( , 2021) (see also references therein) who used Cahn-Hilliard theory to model and simulate the spatio-temporal evolution of the emergent phase separation patterns; the relevant equations for a binary polymer blend are expressed by where ϕ represents the volume fraction of one of the polymers in the blend, M is a constant mobility parameter, and μ is a generalized chemical potential, which can be derived from the variational derivative of the Gibbs free energy functional: here, f denotes the homogeneous contribution to the Gibbs free energy per monomer, which is a nonconvex function of ϕ, and λ is a gradient free energy parameter. Numerical solutions of the above equations are obtained subject to Neumann conditions: ...
... Owoyele et al. [23] used AL to perform simulation-based data generation, ML learning and surrogate optimization to refine solution in the vicinity of predicted optimum parameters for design of a compression ignition engine. Gonçalves et al. [24] studied the generation of simulation-based surrogate models with the task of parameter domain exploration using various sampling and regression-based AL strategies. In a similar work, Pan et al. [25] , used AL for developing surrogate models for industrial fluid flow case studies under a constraint of a limited function evaluations. ...
... Other commonly modelled systems are road infrastructure, atmospheric pollution, sewage, noise pollution, water supply, energy demand and public transport. For example, in the city of London, multiple systems such as the sewage network, electricity supply, energy demand and renewable energy resources have been modelled in two projects carried out by Whyte et al. (2019) andO'Dwyer et al. (2020). The physical 3D model is generally used as the foundation of urban digital twins because of its ease of use and creation compared to other modelled systems. ...
... For example, transmission network managers in Belgium are helping the network absorb more intermittent renewable energy by sharing computer platforms [23]. NDI provides a reliable technological path for building "smart cities" and promoting the coordination of low-carbon energy [24]. The market is also an important way to allocate resources. ...
... Shortly, scientists expect that cities will experience considerable changes as technology and social necessities change; it is this continual change that puts pressure on urban architects and designers to adapt. By 2050, it is projected that between 66 and 70 percent of people on Earth will live in cities [1]. The impact of urbanization on towns, neighborhoods, infrastructure, and the environment has been the subject of several studies [2] regarding the maintenance of dependable, energy-efficient services without compromising the ease and enjoyment of their users. ...