It is a challenge for traditional building control systems to meet occupants’ needs in shared spaces due to the lack of understanding of individual occupant thermal preferences. This is a barrier to balancing energy efficiency and indoor environmental quality (IEQ). Advanced statistical learning methods offer new solutions towards more energy-efficient and user-centric control logics. In this work, a control logic is proposed to optimise the heating, ventilation and air conditioning (HVAC) operation based on thermal comfort archetype preferences, leveraging the ASHRAE Global Thermal Comfort Database II in conjunction with energy simulations. First, we apply the k-means clustering algorithm to categorize occupants into different archetypes regarding their common feedback on the thermal environment. Then, we fit a Bayesian logistic regression model to predict the thermal comfort preferences of different archetypes based on IEQ data. Finally, we identify two occupant-centric control logics to optimize HVAC operation to meet occupants’ requirements: (i) considering a unified response of thermal comfort in the space, and (ii) ensuring the dynamic optimal setpoint when conflicting occupant archetypes are present. Having compared this control logic with a common rule-based logic, our results demonstrate the potential of occupant-centric controls and the importance of multi-objective metrics in accounting for energy efficiency.@en

In this paper, we introduce multi-objective deep centralized multi-agent actor-critic (MO-DCMAC), a multi-objective reinforcement learning method for infrastructural maintenance optimization, an area traditionally dominated by single-objective reinforcement learning (RL) approaches. Previous single-objective RL methods combine multiple objectives, such as probability of collapse and cost, into a singular reward signal through reward-shaping. In contrast, MO-DCMAC can optimize a policy for multiple objectives directly, even when the utility function is nonlinear. We evaluated MO-DCMAC using two utility functions, which use probability of collapse and cost as input. The first utility function is the threshold utility, in which MO-DCMAC should minimize cost so that the probability of collapse is never above the threshold. The second is based on the failure mode, effects, and criticality analysis methodology used by asset managers to assess maintenance plans. We evaluated MO-DCMAC, with both utility functions, in multiple maintenance environments, including ones based on a case study of the historical quay walls of Amsterdam. The performance of MO-DCMAC was compared against multiple rule-based policies based on heuristics currently used for constructing maintenance plans. Our results demonstrate that MO-DCMAC outperforms traditional rule-based policies across various environments and utility functions.@en

Long-term inspection and maintenance (I&M) planning, a multi-stage stochastic optimization problem, can be efficiently formulated as a partially observable Markov decision process (POMDP). However, within this context, single-agent approaches do not scale well for large multi-component systems since the joint state, action and observation spaces grow exponentially with the number of components. To alleviate this curse of dimensionality, cooperative decentralized approaches, known as decentralized POMDPs, are often adopted and solved using multi-agent deep reinforcement learning (MADRL) algorithms. This paper examines the centralization vs. decentralization performance of MADRL formulations in I&M planning of multi-component systems. Towards this, we set up a comprehensive computational experimental program focused on k-out-of-n system configurations, a common and broadly applicable archetype of deteriorating engineering systems, to highlight the manifestations of MADRL strengths and pathologies when optimizing global returns under varying decentralization relaxations.@en

Structural systems must satisfy multiple performance and functionality requirements during their life cycle, withstanding safety-reducing degradation mechanisms and hazards. Intervention strategies must be planned accordingly to maintain structural integrity and minimize total life-cycle costs and risks, posing a complex optimization problem. Recent advances in multi-agent deep reinforcement learning (DRL) in conjunction with partially observable Markov Decision Processes (POMDPs) have shown great potential for determining optimal structural integrity management policies for systems with large state and action spaces compared to traditional decision practices. This paper tackles the maintenance optimization problem of aging bridges in seismic-prone areas, creating an updatable environment that embeds chloride-induced corrosion and state-dependent seismic fragility throughout the bridge life-cycle. The evolution of the environment is captured by a dynamic Bayesian network, and it is further integrated with decentralized multi-agent DRL algorithms to identify near-optimal lifecycle decisions under risk constraints. Results on a multi-component bridge system show the suitability of the developed framework for minimizing expected life-cycle costs, and for providing detailed and adaptive policies that significantly outperform traditional condition- and time-based maintenance plans.@en

Effective transportation network management systems should consider safety and sustainability objectives. Existing research on large-scale transportation network management often employs the assumption that bridges can be considered individually under these objectives. However, this simplification misses accurate system-level representations, induced by multiple components, network topology, and global maintenance actions. To address these limitations, this paper presents a deep reinforcement learning (DRL) framework that draws inspiration from biological learning behaviors to determine optimal life-cycle management policies. It incorporates synergetic branches and hierarchical rewards, factorizing the action space and, thereby, diminishing system complexity from exponential to linear with respect to the number of bridges. Extensive experiments based on a realistic case study demonstrate that the proposed method outperforms expert maintenance strategies and state-of-the-art decision-making methods. Overall, the proposed DRL framework can assist engineers by offering adaptive solutions to maintenance planning. It also provides solutions that address large action spaces within complex systems.

@en

Assessing life-cycle seismic safety of aging reinforced concrete bridges is a challenging engineering task. Deterioration phenomena reduce structural capacity, exacerbating poor design choices that are typical of old bridges, while also being characterized by major uncertainties. Management of engineering systems in highly uncertain environments can be efficiently addressed through Markov decision processes, which rely on dynamic Bayesian networks to model the deteriorating system’s life-cycle. However, there is still a gap in developing virtual environments that can seamlessly fit in such advanced algorithmic decision-making frameworks, especially under life-cycle seismic behavior considerations. In this study, we develop a dynamic Bayesian network capable of incorporating disparate uncertainties related to chloride-induced corrosion and seismic action, aiming at providing fragility curves over the bridge service life. The framework is applied to a prototype bridge encapsulating key risk-prone features. Using a multi-component approach, the developed network provides valuable insights into the fragility evaluation of both the system and individual components. Markovian transitions among component deterioration states are computed by combining corrosion initiation and propagation models with non-stationary Gamma processes. Subsequently, state-dependent fragilities are obtained through probabilistic seismic assessment based on non-linear dynamic analyses and multinomial logistic regression. Results show that the approach sheds light on the risk interplay mechanisms between components and the system, and on how different corrosion scenarios affect the system fragility. Discussion is finally provided on how these risk considerations can be interpreted for decision-making, allowing for better repair and retrofit strategies.@en

Generative Artificial Intelligence (AI) promises to make a vast impact across disciplines, including transforming the architectural design process by autonomously generating full building geometries. One form of generative deep learning that has been used to create 2D and 3D representations of objects is Generative Adversarial Networks (GANs). Existing literature, however, has limited applications that utilize 3D data for building geometry generation, with previous studies focused on low-scale 3D geometries suitable for objects such as chairs or cars. This paper develops a new GAN architecture to produce high-resolution feasible building geometry. The training dataset used is a selection of 3D models of single-family homes from an existing database, pre-processed for the specific application. State-of-the-art GAN models are initially tested to establish baseline performance and applicability potential. Then, a systematic study is performed to identify the structure and hyperparameters necessary to successfully fit a GAN to this design task. The successful architecture, named 3DBuildingGAN, uses a combination of Wasserstein loss with gradient penalty, leaky rectified linear units for neuron activation in the generator and the critic, and the root mean squared propagation optimizer with a fixed learning rate. The proposed model generates outputs similar in size, shape, and proportion to the training data with minimal noise in the output. Evaluation of memorization properties indicates open research directions, such as incorporating memorization rejection and training on larger data sets. Finally, the study reflects on how AI algorithms can reshape creativity through data-driven design solutions.@en

It is now within reach to use generative artificial intelligence (AI) to autonomously generate full building geometries. However, existing literature utilizing 3D data has focused to a limited degree on architecture and engineering disciplines. A critical first step to expanding the use of generative deep learning models in generative design research is making training data available. This study investigates 3D building model data characteristics that make it suitable for generative AI applications. Key data set attributes are identified through a systematic review of the object-containing datasets currently used to train state-of-the-art 3D GANs. These requirements are then compared to attributes of existing available building datasets. This comparison shows that publicly available data sets of 3D building models lack essential characteristics for generative deep learning. Features that make these building models inadequate for the task include but are not limited to, their mesh formats, low resolution and levels of detail, and inclusion of irrelevant geometry. To achieve the desired properties in this work, necessary transformations of the data are incorporated into a tailored preprocessing pipeline. The pipeline is applied to an existing dataset that contains 3D models of single-family homes. The transformed dataset is tested within state-of-the-art GAN models to assess training performance and document future data requirements for applying deep generative design to buildings. Our experiments show promise for the impact that architectural datasets can make on deep learning applications within the discipline. It also highlights the need for additional 3D building model data to increase the diversity and robustness of new designs.@en

Façade engineering is facing an era of extraordinary challenge to meet the surge in demand for buildings that are environmentally sustainable and enhance occupant wellbeing. Facades, also known as building envelopes, play a major role in the resource-efficiency of buildings and the quality of its indoor environment. Consequently, the development of effective design approaches is crucial for generating appropriate façade solutions. Façade design is complex and multi-disciplinary involving several and oftentimes conflicting performance criteria. Systematic and holistic design procedures are, therefore, required to achieve optimal trade-offs. Over the last decades, researchers in this field have used computational tools and power to address this challenging problem within the context of multi-criteria design approaches. This paper reviews the existing research in this field, and presents the state-of-the-art review from simple to advanced decision-making procedures currently used at the early design stages, where decisions have a disproportionally large impact on the façade performance. The paper provides a complete description of the design variables and objectives typically involved. Alternative multi-criteria design methodologies regarding discrete decisions and automated optimization are reviewed, each with salient pros/cons, and overall conclusions are drawn. Finally, the paper discusses ongoing trends and research needs, namely, the development of uncertainty-based procedures to enable more informed decision-making; the inclusion of structural/seismic safety considerations in the design process to achieve higher socio-economic benefits; the integration of smart building information modeling and processing technologies to facilitate smarter design decisions; and the adoption of integrated design approaches to promote climate-adaptive solutions that enhance resilience.

@en

A key computational challenge in maintenance planning for deteriorating structures is to concurrently secure (i) optimality of decisions over long planning horizons, and (ii) accuracy of realtime parameter updates in high-dimensional stochastic spaces. Both are often encumbered by the presence of discretized continuous-state models that describe the underlying deterioration processes, and the emergence of combinatorial decision spaces due to multi-component environments. Recent advances in Deep Reinforcement Learning (DRL) formulations for inspection and maintenance planning provide us with powerful frameworks to handle efficiently near-optimal decision-making in immense state and action spaces without the need for offline system knowledge. Moreover, Bayesian Model Updating (BMU), aided by advanced sampling methods, allows us to address dimensionality and accuracy issues related to discretized degradation processes. Building upon these concepts, we develop a joint framework in this work, coupling DRL, more specifically deep Q-learning and actor-critic algorithms, with BMU through Hamiltonian Monte Carlo. Single- and multi-component systems are examined, and it is shown that the proposed methodology yields reduced lifelong maintenance costs, and policies of high fidelity and sophistication compared to traditional optimized time- and condition-based maintenance strategies.@en

CAAD Futures is a biennial international conference on Computer-Aided Architectural Design under the umbrella of the CAAD Futures Foundation, and it is active world-wide in advancing and documenting related research. On 5–7 July 2023, the 20th CAAD Futures conference was hosted at Delft University of Technology. The CAAD Futures Foundation was established in 1985, holding the first conference on 18–19 September of that year at the very same University. The return of the conference to Delft for its 20thedition offered a chance to reflect on the past, present and future role of Computation in Architecture and the Built Environment. With reference to the theme of “INTERCONNECTIONS: Co-computing beyond boundaries”, CAAD Futures 2023 reflected on the role of computation to interconnect in and for Architectural Design.@en

The application of Deep Reinforcement Learning (DRL) for the management of engineering systems has shown very promising results in terms of optimality and scalability. The interpretability of these policies by decision-makers who are so far mostly familiar with traditional approaches is also needed for implementation. In this work, we address this topic by providing a comprehensive overview of POMDP- and DRL-based management policies, along with simulation-based implementation details, for facilitating their interpretation. By mapping a sufficient statistic, namely a belief state, to the current optimal action, POMDP-DRL strategies are able to automatically adapt in time considering long-term sought objectives and the prior history. Through simulated policy realizations, POMDP-DRL-based strategies identified for representative inspection and maintenance planning settings are thoroughly analyzed. The results reveal that if the decision-maker opts for an alternative, even suboptimal, action other than the one suggested by the DRL-based policy, the belief state will be accordingly updated and can still be used as input for the remainder of the planning horizon, without any requirements for model retraining.@en

Maintenance planning of engineering systems is typically posed as a discrete stochastic optimal control problem, as it refers to determining a series of distinct interventions that upkeep structural integrity. Advanced algorithmic schemes within the joint framework of Partially Observable Markov Decision Processes (POMDPs) and multi-agent Deep Reinforcement Learning (DRL) have been recently able to approximate well global optima for this complex problem, outperforming existing time- and condition-based decision strategies. Integral to their success is the hypothesis that system components represent individual agents who form cooperative policies to minimize a central life-cycle cost. Thereby, the policy output scales linearly with the number of components, alleviating the curse of dimensionality related to combinatorial choices. State complexity and long-term optimality are handled efficiently via deep learning and POMDP principles, respectively. However, the efficiency of multi-agent coordination can fade as the number of agents increases. To this end, we propose a new formulation: we pose the problem as a continuous-control dynamic resource allocation one, combining hierarchical DRL and mixed-integer programming. Moving from flat decentralized to hierarchical multi-agent decompositions allows us to improve further the policy output scalability. The new Adaptive Knapsack Hierarchical Resource Allocator (AK-HRA) DRL architecture distributes available resources within the system, creating local, independently solvable, multi-choice knapsack optimization problems. By design, AK HRA allows decision-makers to inscribe known hierarchical structures and local decision rules in their architectures, thereby enhancing control and interpretability over the solution space. The efficacy of the new approach is demonstrated in a multi-component reliability system subject to stochastic deterioration.@en

This work develops a computational method that produces algorithmically generated design forms, able to overcome inherent challenges related to the use of cast glass for the creation of monolithic structural components with light permeability. Structural Topology Optimization (TO) has a novel applicability potential, as decreased mass is associated with shorter annealing times and, thus, considerably improved manufacturability in terms of time, energy, and cost efficiency. However, realistic TO in such structures is currently hindered by existing mathematical formulations and commercial software capabilities. Incorporating annealing constraints into the optimization problem is an essential feature that needs to be accommodated, whereas the brittle nature of glass invokes asymmetric stress failure criteria that cannot be captured by conventional ductile plasticity surfaces or uniform stress constraints. This paper addresses the approximation problems in the evaluation of principal stresses while concurrently incorporating annealing-related manufacturing constraints into a unified TO formulation. A mass minimization objective is articulated, as this is the most critical factor for cast glass structures. To ensure the structural integrity and manufacturability of the component, the applied constraints refer both to the glass material/structural properties and to criteria that ensue from the annealing and fabrication processes. The developed code is based on the penalized artificial density interpolation scheme and the optimization problem is solved with the interior-point method. The proposed formulation is applied in a planar design domain to explore how different glass compositions and structural design strategies affect the final shape. Upon extraction of the optimized shape, the structural performance of the respective 3D structures is validated with respect to performance constraint violations using the Ansys software. Finally, brief guidelines on the practical aspects of the manufacturing process are provided.@en

User experience and satisfaction with the facade play a significant role in user comfort and energy efficiency of buildings. This paper explores the concept of User-Facade archetypes to inform the user-centred design of shading devices based on the perceived level of importance of different environmental domains at the workplace. A questionnaire was developed to collect data on users’ perceived level of importance of different environmental domains, user characteristics and other preferences. Based on the associated level of importance of the domains affected by shading devices (thermal conditions, access to daylight, access to outdoor view, privacy and glare mitigation), users were then clustered into eight different archetypes, which associated different "weights" to each comfort domain. The study also found a significant correlation between the associated level of importance and the reported frequency of interaction with shadings because of thermal comfort, glare mitigation or privacy. Overall, users that associated high levels of importance to several environmental domains also reported high perceived levels of importance for personal control at the workplace. Only one archetype reported low importance for personal control at the workplace. Further work is required to validate these archetypes by capturing actual user behaviour and preferences in real workplaces. However, these findings provide preliminary and valuable insights into the possibility of clustering users on their preferences and using this for informing a more user-centred design or operation of shading devices.

@en

Bridge structures are exposed to several chronic and abrupt stressors, among which the combined effects of corrosion and earthquakes pose a major threat to their long-term safety. Probabilistic risk assessment frameworks that quantify and propagate uncertainties inherent to these phenomena are necessary to mitigate this threat. This paper proposes a dynamic Bayesian network for state-dependent seismic fragility functions, capturing corrosion and seismic effects over time. Markovian transitions among deterioration states for different bridge components are developed, combining chloride diffusion and corrosion propagation models with non-stationary Gamma processes. State-dependent fragility curves are derived based on non-linear dynamic time-history analyses given possible degradation configurations of the structure, accounting for uncertainties in material, geometry, and deterioration parameters. Record-to-record variability is captured using synthetic ground motions. Results on a 4-span Gerber bridge showcase the suitability of the framework for describing life-cycle fragility, and its capacity for embedding in advanced algorithmic decision-making workflows is discussed.@en

To preserve structural safety of deteriorating engineering systems through optimal maintenance, it is imperative to efficiently integrate structural health information with decision-making optimization frameworks. Although there may be abundance of available data, these are often uncertain and incomplete. In addition, joint inspection and maintenance (I&M) optimization is inherently complex due to high-dimensional state and action spaces, stochastic objectives, long planning horizons, and various constraints, among others. As shown recently, these computational challenges can be effectively addressed through optimization principles of Partially Observable Markov Decision Processes (POMDPs) and constrained Deep Reinforcement Learning (DRL). The POMDP framework provides a way of updating the decision-maker's perception about the system state by naturally incorporating the Value of Information (VoI) in the optimality equations. As such, optimal observation-gathering actions are those which guide maintenance decisions towards reduced life-cycle costs and risks. The role of VoI in DRL-driven I&M has also been shown to be central to the formation of policy gradients, which are necessary to obtain the optimal I&M plan with deep learning actor-critic architectures. Leveraging this property, a recently devised DRL architecture is further examined in this work, consisting of fully decoupled 'maintainer' and 'inspector' actors, which allow for greater efficacy and interpretability in multi-agent DRL settings. Several numerical analyses are carried out to assess the performance of the relevant architectures on stochastic systems with a varying number of components, multiple maintenance-inspection actions per component, and system-level failure risks.@en

Long-term inspection and maintenance (I&M) planning, a multi-stage stochastic optimization problem, can be efficiently formulated as a partially observable Markov decision process (POMDP). How- ever, within this context, single-agent approaches do not scale well for large multi-component systems since the joint state, action and observation spaces grow exponentially with the number of components. To alleviate this curse of dimensionality, cooperative decentralized approaches, known as decentralized POMDPs, are often adopted and solved using multi-agent deep reinforcement learning (MADRL) algorithms. This paper examines the centralization vs. decentralization performance of MADRL formulations in I&M planning of multi-component systems. Towards this, we set up a comprehensive computational experimental program focused on k-out-of-n system configurations, a common and broadly applicable archetype of deteriorating engineering systems, to highlight the manifestations of MADRL strengths and pathologies when optimizing global returns under varying decentralization relaxations in such systems.
@en

In the context of modern engineering, environmental, and societal concerns, there is an increasing demand for methods able to identify rational management strategies for civil engineering systems, minimizing structural failure risks while optimally planning inspection and maintenance (I&M) processes. Most available methods simplify the I&M decision problem to the component level, often assuming statistical, structural, or cost independence among components, due to the computational complexity associated with global optimization methodologies under joint system-level state descriptions. In this paper, we propose an efficient algorithmic framework for inference and decision-making under uncertainty for engineering systems exposed to deteriorating environments, providing optimal management strategies directly at the system level. In our approach, the decision problem is formulated as a factored partially observable Markov decision process, whose dynamics are encoded in Bayesian network conditional structures. The methodology can handle environments under equal or general, unequal deterioration correlations among components, through Gaussian hierarchical structures and dynamic Bayesian networks, decoupling the originally joint system state space to component networks conditional on shared random variables. In terms of policy optimization, we adopt a deep decentralized multi-agent actor-critic (DDMAC) reinforcement learning approach, in which the policies are approximated by actor neural networks guided by a critic network. By including deterioration dependence in the simulated environment, and by formulating the cost model at the system level, DDMAC policies intrinsically consider the underlying system-effects. This is demonstrated through numerical experiments conducted for both a 9-out-of-10 system and a steel frame under fatigue deterioration. Results demonstrate that DDMAC policies offer substantial benefits when compared to state-of-the-art heuristic approaches. The inherent consideration of system-effects by DDMAC strategies is also interpreted based on the learned policies.

@en

Fragility analysis aims to compute the probabilities of a system exceeding certain damage conditions given different levels of hazard intensity. Fragility analysis is therefore a key process of performance-based earthquake engineering, with a number of approaches developed and widely recognized, including Incremental Dynamic Analysis (IDA), Multiple Stripe Analysis (MSA), and cloud analysis. Additionally, extended fragility analysis has recently been shown to possess important attributes of mathematical consistency and extensibility. This work provides a critical review of the different fragility methods by explaining the underlying probabilistic models and assumptions, as well as their connections to the extended fragility method. It is proven that IDA-based fragility curves provide an upper bound of the actual fragility, and cloud analysis manifests suboptimality issues arising from its underlying assumptions. MSA is identified to be a probit-linked Bernoulli regression model, similar to the one proposed by Shinozuka and coworkers. The latter, in turn, is shown to be a limiting subcase of the generalized linear model framework introduced within the extended fragility analysis. The paper first presents a simple case of one intensity measure and two damage condition states, and the discussion is subsequently extended to more general cases of multiple intensity measures and damage states. The discussed attributes are demonstrated in several numerical applications. Overall, this work aims to provide new insights on fragility methods, enabling efficient, accurate, and consistent estimations of structural performance, as well as promoting new research directions in earthquake engineering and other related fields.@en