Urban communities rely on built utility infrastructures as critical lifelines that provide essential services such as water, gas, and power, to sustain modern socioeconomic systems. These infrastructures consist of underground and surface-level assets that are operated and geo-distributed over large regions where continuous monitoring for anomalies is required but challenging to implement. This article addresses the problem of deploying heterogeneous Internet of Things sensors in these networks to support future decision-support tasks, for example, anomaly detection, source identification, and mitigation. We use stormwater as a driving use case; these systems are responsible for drainage and flood control, but act as conduits that can carry contaminants to the receiving waters. Challenges toward effective monitoring include the transient and random nature of the pollution incidents, the scarcity of historical data, the complexity of the system, and technological limitations for real-time monitoring. We design a SemanTics-aware sEnsor Placement framework (STEP) to capture pollution incidents using structural, behavioral, and semantic aspects of the infrastructure. We leverage historical data to inform our system with new, credible instances of potential anomalies. Several key topological and empirical network properties are used in proposing candidate deployments that optimize the balance between multiple objectives. We also explore the quality of anomaly representation in the network through new perspectives, and provide techniques to enhance the realism of the anomalies considered in a network. We evaluate STEP on six real-world stormwater networks in Southern California, USA, which shows its efficacy in monitoring areas of interest over other baseline methods.

Many documents are produced over the years of managing assets, particularly those with long lifespans. However, during this time, the assets may deviate from their original as-designed or as-built state. This presents a significant challenge for tasks that occur in later life phases but require precise knowledge of the asset, such as retrofit, where the assets are equipped with new components. For a third party who is neither the original manufacturer nor the operator, obtaining a comprehensive understanding of the asset can be a tedious process, as this requires going through all available but often fragmented information and documents. While common knowledge regarding the domain or general type of asset can be helpful, it is often based on the experiences of engineers and is, therefore, only implicitly available. This article presents a graph-based information management system that complements traditional PLM systems and helps connect fragments by utilizing generic information about assets. To achieve this, techniques from systems engineering and data science are used. The overarching management platform also includes geometric analyses and operations that can be performed with geometric and product information extracted from STEP files. While the management itself is first described generically, it is also later applied to cabin retrofit in aviation. A mock-up of an Airbus A320 is utilized as the case study to demonstrate further how the platform can provide benefits for retrofitting such long-living assets.

We propose a physics-constrained convolutional neural network (PC-CNN) to solve two types of inverse problems in partial differential equations (PDEs), which are nonlinear and vary both in space and time. In the first inverse problem, we are given data that is offset by spatially varying systematic error (i.e., the bias, also known as the epistemic uncertainty). The task is to uncover the true state, which is the solution of the PDE, from the biased data. In the second inverse problem, we are given sparse information on the solution of a PDE. The task is to reconstruct the solution in space with high resolution. First, we present the PC-CNN, which constrains the PDE with a time-windowing scheme to handle sequential data. Second, we analyze the performance of the PC-CNN to uncover solutions from biased data. We analyze both linear and nonlinear convection-diffusion equations, and the Navier–Stokes equations, which govern the spatiotemporally chaotic dynamics of turbulent flows. We find that the PC-CNN correctly recovers the true solution for a variety of biases, which are parameterized as non-convex functions. Third, we analyze the performance of the PC-CNN for reconstructing solutions from sparse information for the turbulent flow. We reconstruct the spatiotemporal chaotic solution on a high-resolution grid from only 1% of the information contained in it. For both tasks, we further analyze the Navier–Stokes solutions. We find that the inferred solutions have a physical spectral energy content, whereas traditional methods, such as interpolation, do not. This work opens opportunities for solving inverse problems with partial differential equations.

Transfer learning has been highlighted as a promising framework to increase the accuracy of the data-driven model in the case of data sparsity, specifically by leveraging pretrained knowledge to the training of the target model. The objective of this study is to evaluate whether the number of requisite training samples can be reduced with the use of various transfer learning models for predicting, for example, the chemical source terms of the data-driven reduced-order modeling (ROM) that represents the homogeneous ignition of a hydrogen/air mixture. Principal component analysis is applied to reduce the dimensionality of the hydrogen/air mixture in composition space. Artificial neural networks (ANNs) are used to regress the reaction rates of principal components, and subsequently, a system of ordinary differential equations is solved. As the number of training samples decreases in the target task, the ROM fails to predict the ignition evolution of a hydrogen/air mixture. Three transfer learning strategies are then applied to the training of the ANN model with a sparse dataset. The performance of the ROM with a sparse dataset is remarkably enhanced if the training of the ANN model is restricted by a regularization term that controls the degree of knowledge transfer from source to target tasks. To this end, a novel transfer learning method is introduced, Parameter control via Partial Initialization and Regularization (PaPIR), whereby the amount of knowledge transferred is systemically adjusted in terms of the initialization and regularization schemes of the ANN model in the target task.

The performance and confidence in fault detection and diagnostic systems can be undermined by data pipelines that feature multiple compounding sources of uncertainty. These issues further inhibit the deployment of data-based analytics in industry, where variable data quality and lack of confidence in model outputs are already barriers to their adoption. The methodology proposed in this paper supports trustworthy data pipeline design and leverages knowledge gained from one fully-observed data pipeline to a similar, under-observed case. The transfer of uncertainties provides insight into uncertainty drivers without repeating the computational or cost overhead of fully redesigning the pipeline. A SHAP-based human-readable explainable AI (XAI) framework was used to rank and explain the impact of each choice in a data pipeline, allowing the decoupling of positive and negative performance drivers to facilitate the successful selection of highly-performing pipelines. This empirical approach is demonstrated in bearing fault classification case studies using well-understood open-source data.

Buildings employ an ensemble of technical systems like those for heating and ventilation. Ontologies such as Brick, IFC, SSN/SOSA, and SAREF have been created to describe such technical systems in a machine-understandable manner. However, these focus on describing system topology, whereas several relevant use cases (e.g., automated fault detection and diagnostics (AFDD)) also need knowledge about the physical processes. While mathematical simulation can be used to model physical processes, these are practically expensive to run and are not integrated with mainstream technical systems ontologies today. We propose to describe the effect of component actuation on underlying physical mechanisms within component stereotypes. These stereotypes are linked to actual component instances in the technical system description, thereby accomplishing an integration of knowledge about system structure and physical processes. We contribute an ontology for such stereotypes and show that it covers 100% of Brick heating, ventilation, and air-conditioning (HVAC) components. We further show that the ontology enables automatically inferring relationships between components in a real-world building in most cases, except in two situations where component dependencies are underreported. This is due to missing component models for passive parts like splits and join in ducts, and hence points at concrete future extensions of the Brick ontology. Finally, we demonstrate how AFDD applications can utilize the resulting knowledge graph to find expected consequences of an action, or conversely, to identify components that may be responsible for an observed state of the process.

Gas furnaces are the prevalent heating systems in Europe, but efforts to decarbonize the energy sector advocate for their replacement with heat pumps. However, this transition poses challenges for power grids due to increased electricity consumption. Estimating this consumption relies on the seasonal performance factor (SPF) of heat pumps, a metric that is complex to model and hard to measure accurately. We propose using an unpaired dataset of smart meter data at the building level to model the heat consumption and the SPF. We compare the distributions of the annual gas and heat pump electricity consumption by applying either the Jensen–Shannon Divergence or the Kolmogorov–Smirnov test. Through evaluation of a real-world dataset, we prove the ability of the methodology to predict the electricity consumption of future heat pumps replacing existing gas furnaces with a focus on single- and two-family buildings. Our results indicate anticipated SPFs ranging between 2.8 and 3.4, based on the Kolmogorov–Smirnov test. However, it is essential to note that the analysis reveals challenges associated with interpreting results when there are single-sided shifts in the input data, such as those induced by external factors like the European gas crisis in 2022. In summary, this extended version of a conference paper shows the viability of utilizing smart meter data to model heat consumption and seasonal performance factor for future retrofitted heat pumps.

Turbulent flows are chaotic and multi-scale dynamical systems, which have large numbers of degrees of freedom. Turbulent flows, however, can be modeled with a smaller number of degrees of freedom when using an appropriate coordinate system, which is the goal of dimensionality reduction via nonlinear autoencoders. Autoencoders are expressive tools, but they are difficult to interpret. This article aims to propose a method to aid the interpretability of autoencoders. First, we introduce the decoder decomposition, a post-processing method to connect the latent variables to the coherent structures of flows. Second, we apply the decoder decomposition to analyze the latent space of synthetic data of a two-dimensional unsteady wake past a cylinder. We find that the dimension of latent space has a significant impact on the interpretability of autoencoders. We identify the physical and spurious latent variables. Third, we apply the decoder decomposition to the latent space of wind-tunnel experimental data of a three-dimensional turbulent wake past a bluff body. We show that the reconstruction error is a function of both the latent space dimension and the decoder size, which are correlated. Finally, we apply the decoder decomposition to rank and select latent variables based on the coherent structures that they represent. This is useful to filter unwanted or spurious latent variables or to pinpoint specific coherent structures of interest. The ability to rank and select latent variables will help users design and interpret nonlinear autoencoders.

Data irregularities, namely small disjuncts, class skew, imbalance, and outliers significantly affect the performance of classifiers. Another challenge posed to classifiers is when new unlabelled data have different characteristics than the training data; this change is termed as a data shift. In this paper, we focus on identifying small disjuncts and dataset shift using the supervised classifier, sequential ellipsoidal partitioning classifier (SEP-C). This method iteratively partitions the dataset into minimum-volume ellipsoids that contain points of the same label, based on the idea of Reduced Convex Hulls. By allowing an ellipsoid that contains points of one label to contain a few points of the other, such small disjuncts may be identified. Similarly, if new points are accommodated only by expanding one or more of the ellipsoids, then shifts in data can be identified. Small disjuncts are distribution-based irregularities that may be considered as being rare but more error-prone than large disjuncts. Eliminating small disjuncts by removal or pruning is seen to affect the learning of the classifier adversely. Dataset shifts have been identified using Bayesian methods, use of confidence scores, and thresholds—these require prior knowledge of the distributions or heuristics. SEP-C is agnostic of the underlying data distributions, uses a single hyperparameter, and as ellipsoidal partitions are generated, well-known statistical tests can be performed to detect shifts in data; it is also applicable as a supervised classifier when the datasets are highly skewed and imbalanced. We demonstrate the performance of SEP-C with UCI, MNIST handwritten digit image, and synthetically generated datasets.

In developing countries, a significant amount of natural gas is used for household water heating, accounting for roughly 50% of total usage. Legacy systems, typified by large water heaters, operate inefficiently by continuously maintaining a large volume of water at a constant temperature, irrespective of demand. With dwindling domestic gas reserves and rising demand, this increases dependence on expensive energy imports.

We introduce a novel Internet of Things (IoT)-inspired solution to understand and predict water usage patterns and only activate the water heater when there’s a predicted demand. This retrofit system is maintenance-free and uses a rechargeable battery powered by a thermoelectric generator (TEG), which capitalizes on the temperature difference between the heater and its environment for electricity. Our study shows a notable 70% reduction in natural gas consumption compared to traditional systems. Our solution offers a sustainable and efficient method for water heating, addressing the challenges of depleting gas reserves and rising energy costs.

Many physical systems exhibit limit-cycle oscillations that can typically be modeled as stochastically driven self-oscillators. In this work, we focus on a self-oscillator model where the nonlinearity is on the damping term. In various applications, it is crucial to determine the nonlinear damping term and the noise intensity of the driving force. This article presents a novel approach that employs a deep operator network (DeepONet) for parameter identification of self-oscillators. We build our work upon a system identification methodology based on the adjoint Fokker–Planck formulation, which is robust to the finite sampling interval effects. We employ DeepONet as a surrogate model for the operator that maps the first Kramers–Moyal (KM) coefficient to the first and second finite-time KM coefficients. The proposed approach can directly predict the finite-time KM coefficients, eliminating the intermediate computation of the solution field of the adjoint Fokker–Planck equation. Additionally, the differentiability of the neural network readily facilitates the use of gradient-based optimizers, further accelerating the identification process. The numerical experiments demonstrate that the proposed methodology can recover desired parameters with a significant reduction in time while maintaining an accuracy comparable to that of the classical finite-difference approach. The low computational time of the forward path enables Bayesian inference of the parameters. Metropolis-adjusted Langevin algorithm is employed to obtain the posterior distribution of the parameters. The proposed method is validated against numerical simulations and experimental data obtained from a linearly unstable turbulent combustor.

Assisted and automated driving functions will rely on machine learning algorithms, given their ability to cope with real-world variations, e.g. vehicles of different shapes, positions, colors, and so forth. Supervised learning needs annotated datasets, and several automotive datasets are available. However, these datasets are tremendous in volume, and labeling accuracy and quality can vary across different datasets and within dataset frames. Accurate and appropriate ground truth is especially important for automotive, as “incomplete” or “incorrect” learning can negatively impact vehicle safety when these neural networks are deployed. This work investigates the ground truth quality of widely adopted automotive datasets, including a detailed analysis of KITTI MoSeg. According to the identified and classified errors in the annotations of different automotive datasets, this article provides three different criteria collections for producing improved annotations. These criteria are enforceable and applicable to a wide variety of datasets. The three annotations sets are created to (i) remove dubious cases; (ii) annotate to the best of human visual system; and (iii) remove clear erroneous BBs. KITTI MoSeg has been reannotated three times according to the specified criteria, and three state-of-the-art deep neural network object detectors are used to evaluate them. The results clearly show that network performance is affected by ground truth variations, and removing clear errors is beneficial for predicting real-world objects only for some networks. The relabeled datasets still present some cases with “arbitrary”/“controversial” annotations, and therefore, this work concludes with some guidelines related to dataset annotation, metadata/sublabels, and specific automotive use cases.

The design of gas turbine combustors for optimal operation at different power ratings is a multifaceted engineering task, as it requires the consideration of several objectives that must be evaluated under different test conditions. We address this challenge by presenting a data-driven approach that uses multiple probabilistic surrogate models derived from Gaussian process regression to automatically select optimal combustor designs from a large parameter space, requiring only a few experimental data points. We present two strategies for surrogate model training that differ in terms of required experimental and computational efforts. Depending on the measurement time and cost for a target, one of the strategies may be preferred. We apply the methodology to train three surrogate models under operating conditions where the corresponding design objectives are critical: reduction of NOx emissions, prevention of lean flame extinction, and mitigation of thermoacoustic oscillations. Once trained, the models can be flexibly used for different forms of a posteriori design optimization, as we demonstrate in this study.

Current fault diagnosis (FD) methods for heating, ventilation, and air conditioning (HVAC) systems do not accommodate for system reconfigurations throughout the systems’ lifetime. However, system reconfiguration can change the causal relationship between faults and symptoms, which leads to a drop in FD accuracy. In this paper, we present Fault-Symptom Brick (FSBrick), an extension to the Brick metadata schema intended to represent information necessary to propagate system configuration changes onto FD algorithms, and ultimately revise FSRs. We motivate the need to represent FSRs by illustrating their changes when the system reconfigures. Then, we survey FD methods’ representation needs and compare them against existing information modeling efforts within and outside of the HVAC sector. We introduce the FSBrick architecture and discuss which extensions are added to represent FSRs. To evaluate the coverage of FSBrick, we implement FSBrick on (i) the motivational case study scenario, (ii) Building Automation Systems’ representation of FSRs from 3 HVACs, and (iii) FSRs from 12 FD method papers, and find that FSBrick can represent 88.2% of fault behaviors, 92.8% of fault severities, 67.9% of symptoms, and 100% of grouped symptoms, FSRs, and probabilities associated with FSRs. The analyses show that both Brick and FSBrick should be expanded further to cover HVAC component information and mathematical and logical statements to formulate FSRs in real life. As there is currently no generic and extensible information model to represent FSRs in commercial buildings, FSBrick paves the way to future extensions that would aid the automated revision of FSRs upon system reconfiguration.

Alarm flood classification (AFC) methods are crucial in assisting human operators to identify and mitigate the overwhelming occurrences of alarm floods in industrial process plants, a challenge exacerbated by the complexity and data-intensive nature of modern process control systems. These alarm floods can significantly impair situational awareness and hinder decision-making. Existing AFC methods face difficulties in dealing with the inherent ambiguity in alarm sequences and the task of identifying novel, previously unobserved alarm floods. As a result, they often fail to accurately classify alarm floods. Addressing these significant limitations, this paper introduces a novel three-tier AFC method that uses alarm time series as input. In the transformation stage, alarm floods are subjected to an ensemble of convolutional kernel-based transformations (MultiRocket) to extract their characteristic dynamic properties, which are then fed into the classification stage, where a linear ridge regression classifier ensemble is used to identify recurring alarm floods. In the final novelty detection stage, the local outlier probability (LoOP) is used to determine a confidence measure of whether the classified alarm flood truly belongs to a known or previously unobserved class. Our method has been thoroughly validated using a publicly available dataset based on the Tennessee-Eastman process. The results show that our method outperforms two naive baselines and four existing AFC methods from the literature in terms of overall classification performance as well as the ability to optimize the balance between accurately identifying alarm floods from known classes and detecting alarm flood classes that have not been observed before.

The intersection of physics and machine learning has given rise to the physics-enhanced machine learning (PEML) paradigm, aiming to improve the capabilities and reduce the individual shortcomings of data- or physics-only methods. In this paper, the spectrum of PEML methods, expressed across the defining axes of physics and data, is discussed by engaging in a comprehensive exploration of its characteristics, usage, and motivations. In doing so, we present a survey of recent applications and developments of PEML techniques, revealing the potency of PEML in addressing complex challenges. We further demonstrate the application of select such schemes on the simple working example of a single degree-of-freedom Duffing oscillator, which allows to highlight the individual characteristics and motivations of different “genres” of PEML approaches. To promote collaboration and transparency, and to provide practical examples for the reader, the code generating these working examples is provided alongside this paper. As a foundational contribution, this paper underscores the significance of PEML in pushing the boundaries of scientific and engineering research, underpinned by the synergy of physical insights and machine learning capabilities.

With global wind energy capacity ramping up, accurately predicting damage equivalent loads (DELs) and fatigue across wind turbine populations is critical, not only for ensuring the longevity of existing wind farms but also for the design of new farms. However, the estimation of such quantities of interests is hampered by the inherent complexity in modeling critical underlying processes, such as the aerodynamic wake interactions between turbines that increase mechanical stress and reduce useful lifetime. While high-fidelity computational fluid dynamics and aeroelastic models can capture these effects, their computational requirements limits real-world usage. Recently, fast machine learning-based surrogates which emulate more complex simulations have emerged as a promising solution. Yet, most surrogates are task-specific and lack flexibility for varying turbine layouts and types. This study explores the use of graph neural networks (GNNs) to create a robust, generalizable flow and DEL prediction platform. By conceptualizing wind turbine populations as graphs, GNNs effectively capture farm layout-dependent relational data, allowing extrapolation to novel configurations. We train a GNN surrogate on a large database of PyWake simulations of random wind farm layouts to learn basic wake physics, then fine-tune the model on limited data for a specific unseen layout simulated in HAWC2Farm for accurate adapted predictions. This transfer learning approach circumvents data scarcity limitations and leverages fundamental physics knowledge from the source low-resolution data. The proposed platform aims to match simulator accuracy, while enabling efficient adaptation to new higher-fidelity domains, providing a flexible blueprint for wake load forecasting across varying farm configurations.

In single-zone multi-node systems (SZMNSs), temperature controls rely on a single probe near the thermostat, resulting in temperature discrepancies that cause thermal discomfort and energy waste. Augmenting smart thermostats (STs) with per-room sensors has gained acceptance by major ST manufacturers. This paper leverages additional sensory information to empirically characterize the services provided by buildings, including thermal comfort, energy efficiency, and demand response (DR). Utilizing room-level time-series data from 1000 houses, metadata from 110,000 houses across the United States, and data from two real-world testbeds, we examine the limitations of SZMNSs and explore the potential of remote sensors. We discover that comfortable DR durations (CDRDs) for rooms are typically 70% longer or 40% shorter than for the room with the thermostat. When averaging, rooms at the control temperature’s bounds are typically deviated around −3 °F to 2.5 °F from the average. Moreover, in 95% of houses, we identified rooms experiencing notably higher solar gains compared to the rest of the rooms, while 85% and 70% of houses demonstrated lower heat input and poor insulation, respectively. Lastly, it became evident that the consumption of cooling energy escalates with the increase in the number of sensors, whereas heating usage experiences fluctuations ranging from −19% to +25%. This study serves as a benchmark for assessing the thermal comfort and DR services in the existing housing stock, while also highlighting the energy efficiency impacts of sensing technologies. Our approach sets the stage for more granular, precise control strategies of SZMNSs.

To maximize its value, the design, development and implementation of structural health monitoring (SHM) should focus on its role in facilitating decision support. In this position paper, we offer perspectives on the synergy between SHM and decision-making. We propose a classification of SHM use cases aligning with various dimensions that are closely linked to the respective decision contexts. The types of decisions that have to be supported by the SHM system within these settings are discussed along with the corresponding challenges. We provide an overview of different classes of models that are required for integrating SHM in the decision-making process to support the operation and maintenance of structures and infrastructure systems. Fundamental decision-theoretic principles and state-of-the-art methods for optimizing maintenance and operational decision-making under uncertainty are briefly discussed. Finally, we offer a viewpoint on the appropriate course of action for quantifying, validating, and maximizing the added value generated by SHM. This work aspires to synthesize the different perspectives of the SHM, Prognostic Health Management, and reliability communities, and provide directions to researchers and practitioners working towards more pervasive monitoring-based decision-support.

We introduce a novel human-centric deep reinforcement learning recommender system designed to co-optimize energy consumption, thermal comfort, and air quality in commercial buildings. Existing approaches typically optimize these objectives separately or focus solely on controlling energy-consuming building resources without directly engaging occupants. We develop a deep reinforcement learning architecture based on multitask learning with humans-in-the-loop and demonstrate how it can jointly learn energy savings, comfort, and air quality improvements for different building and occupant actions. In addition to controlling typical building resources (e.g., thermostat setpoint), our system provides real-time actionable recommendations that occupants can take (e.g., move to a new location) to co-optimize energy, comfort, and air quality. Through real deployments across multiple commercial buildings, we show that our multitask deep reinforcement learning recommender system has the potential to reduce energy consumption by up to 8% in energy-focused optimization, improve all objectives by 5–10% in joint optimization, and improve thermal comfort by up to 21% in comfort and air quality-focused optimization compared to existing solutions.