Bridge infrastructure managers are facing multiple challenges to improve the availability and serviceability of ageing infrastructure, while maintenance planning is constrained by budget restrictions. Many research efforts ongoing, for last few decades, ranging from development bridge management system, decision support tools, optimisation models, life cycle cost analysis, etc. Since transport infrastructures deeply embedded in society, they not only subject technical requirements, but...

We investigate the capabilities of transfer learning in area structural health monitoring. In particular, we are interested damage detection for concrete structures. Typical image datasets such problems relatively small, calling learned representation from a related large-scale dataset. Past efforts using images have mainly considered cross-domain approaches pre-trained IMAGENET models that subsequently fine-tuned target task. However, there rising concerns about generalizability...

Abstract In this paper, we introduce multi-objective deep centralized multi-agent actor-critic (MO-DCMAC), a reinforcement learning method for infrastructural maintenance optimization, an area traditionally dominated by single-objective (RL) approaches. Previous RL methods combine multiple objectives, such as probability of collapse and cost, into singular reward signal through reward-shaping. contrast, MO-DCMAC can optimize policy objectives directly, even when the utility function is...

Data-driven decision support can substantially aid in smart and efficient maintenance planning of road bridges. However, many infrastructure managers primary rely on information obtained during visual inspection to subjectively decide the follow-up actions. The subjective approach is likely lack appropriate use data does not promise cost-effective plans. In this paper, we show that historical operational data, readily available at agencies, vital importance be used effectively for...

Abstract This paper introduces a comprehensive framework for the development of optimal multi-year maintenance plans large number bridges. A plan is said to be when, within given budget, maximum bridges can maintained in best possible year, achieving performance with minimum socio-economic impact. The incorporates heuristic rules, multi-attribute utility theory, discrete Markov chain process, and genetic algorithms find an balance between limited budgets requirements. applicability proposed...

Predictive business process monitoring focuses on predicting future characteristics of a running using event logs. The foresight into execution promises great potentials for efficient operations, better resource management, and effective customer services. Deep learning-based approaches have been widely adopted in mining to address the limitations classical algorithms solving multiple problems, especially next remaining-time prediction tasks. Nevertheless, designing deep neural architecture...

Abstract Cost-effective asset management is an area of interest across several industries. Specifically, this paper develops a deep reinforcement learning (DRL) solution to automatically determine optimal rehabilitation policy for continuously deteriorating water pipes. We approach the problem planning in online and offline DRL setting. In DRL, agent interacts with simulated environment multiple pipes distinct lengths, materials, failure rate characteristics. train using Q-learning (DQN)...

The Adaptive Large Neighborhood Search (ALNS) algorithm has shown considerable success in solving combinatorial optimization problems (COPs). Nonetheless, the performance of ALNS relies on proper configuration its selection and acceptance parameters, which is known to be a complex resource-intensive task. To address this, we introduce Deep Reinforcement Learning (DRL) based approach called DR-ALNS that selects operators, adjusts controls criterion throughout search. proposed method aims...

Aged earthworks constitute a major proportion of European rail infrastructures, the replacement and remediation which poses serious problem. Considering scale networks involved, it is infeasible both in terms track downtime money to replace all these assets. It is, therefore, imperative develop rational means managing slope infrastructure determine best use available resources plan maintenance order criticality. To do so, necessary not just consider structural performance asset but also...

In this study we evaluate 32 unsupervised anomaly detection algorithms on 52 real-world multivariate tabular datasets, performing the largest comparison of to date. On collection $k$-thNN (distance $k$-nearest neighbor) algorithm significantly outperforms most other algorithms. Visualizing and then clustering relative performance considered all identify two clear clusters: one with ``local'' another ``global'' datasets. ``Local'' anomalies occupy a region low density when compared nearby...

Management of structured business processes is interest to both academia and industry, where focuses on the development methods techniques while industry supporting tools.With shift from routine knowledge work, relevance management Unstructured Business Processes (UBP) increasing.However, currently available modeling notations are not optimally suited for UBP.By means a representative example, we investigate limitations Process Model Notation (BPMN) Case (CMMN) in this respect.We derive set...

In bridge management systems, multi-objective decision-making has emerged as a decision support technique to integrate various technical information and stakeholder values.Different multicriteria making techniques tools have been developed in the last three decades.This paper presents an overview of different approaches at object network level, with purpose incorporating aspects performance goals, which may vary according technical, environmental, economic social factors.The example...

Out-of-distribution (OOD) detection is concerned with identifying data points that do not belong to the same distribution as model's training data. For safe deployment of predictive models in a real-world environment, it critical avoid making confident predictions on OOD inputs can lead potentially dangerous consequences. However, largely remains an under-explored area audio (and speech) domain. This despite fact central modality for many tasks, such speaker diarization, automatic speech...

Large-scale infrastructure systems are crucial for societal welfare, and their effective management requires strategic forecasting intervention methods that account various complexities. Our study addresses two challenges within the Prognostics Health Management (PHM) framework applied to sewer assets: modeling pipe degradation across severity levels developing maintenance policies. We employ Multi-State Degradation Models (MSDM) represent stochastic process in pipes use Deep Reinforcement...

The Job Shop Scheduling Problem (JSSP) is a complex combinatorial optimization problem. There has been growing interest in using online Reinforcement Learning (RL) for JSSP. While RL can quickly find acceptable solutions, especially larger problems, it produces lower-quality results than traditional methods like Constraint Programming (CP). A significant downside of that cannot learn from existing data, such as solutions generated CP, requiring them to train scratch, leading sample...

Optimal maintenance is one of the key concerns for asset-intensive industries in terms reducing downtime and occurring costs. The advancement data-driven technologies, affordable computing powers, growing amounts data introduced a paradigm with name predictive (PdM). PdM seeks to find out an optimal moment asset, where no early intervention leads undue extra cost, late activity poses safety risk. With instrumentation cyber-physical system on assets, transforms typical structure into smart...

We introduce an open-source GitHub repository containing comprehensive benchmarks for a wide range of machine scheduling problems, including Job Shop Scheduling (JSP), Flow (FSP), Flexible (FJSP), FJSP with Assembly constraints (FAJSP), Sequence-Dependent Setup Times (FJSP-SDST), and the online (with job arrivals). Our primary goal is to provide centralized hub researchers, practitioners, enthusiasts interested in tackling challenges.