As LLM-based multi-agent systems (MAS) become more autonomous, their free-form interactions increasingly dominate system behavior. However, scaling the number of agents often amplifies context pressure, coordination errors, and system drift. It is well known that building robust MAS requires more than prompt tuning or increased model intelligence. It necessitates engineering discipline focused on architecture to manage complexity under uncertainty. We characterize agentic software by a core property: \emph{runtime generation and evolution under uncertainty}. Drawing upon and extending software engineering experience, especially object-oriented programming, this paper introduces \emph{Loosely-Structured Software (LSS)}, a new class of software systems that shifts the engineering focus from constructing deterministic logic to managing the runtime entropy generated by View-constructed programming, semantic-driven self-organization, and endogenous evolution. To make this entropy governable, we introduce design principles under a three-layer engineering framework: \emph{View/Context Engineering} to manage the execution environment and maintain task-relevant Views, \emph{Structure Engineering} to organize dynamic binding over artifacts and agents, and \emph{Evolution Engineering} to govern the lifecycle of self-rewriting artifacts. Building on this framework, we develop LSS design patterns as semantic control blocks that stabilize fluid, inference-mediated interactions while preserving agent adaptability. Together, these abstractions improve the \emph{designability}, \emph{scalability}, and \emph{evolvability} of agentic infrastructure. We provide basic experimental validation of key mechanisms, demonstrating the effectiveness of LSS.
Carbon credit systems have emerged as a policy tool to incentivize emission reductions and support the transition to clean energy. Reliable carbon-credit certification depends on mechanisms that connect actual, measured renewable-energy production to verifiable emission-reduction records. Although blockchain and IoT technologies have been applied to emission monitoring and trading, existing work offers limited support for certification processes, particularly for small and medium-scale renewable installations. This paper introduces a blockchain-based carbon-credit certification architecture, demonstrated through a 100 kWp photovoltaic case study, that integrates real-time IoT data collection, edge-level aggregation, and secure on-chain storage on a permissioned blockchain with smart contracts. Unlike approaches focused on trading mechanisms, the proposed system aligns with European legislation and voluntary carbon-market standards, clarifying the practical requirements and constraints that apply to photovoltaic operators. The resulting architecture provides a structured pathway for generating verifiable carbon-credit records and supporting third-party verification.
Sergey I. Abrakhin, Anastasiya V. Lukina, Mikhail S. Lisyatnikov
et al.
Estimating the load-bearing capacity and predicting the residual strength of existing structures is one of the most difficult tasks. Such prediction is usually performed on the basis of experimental destructive testing of samples. A methodology for predicting the residual strength of wooden structures is proposed, based on the results of experimental studies to determine the short-term resistance of pure wood. Wooden rafter systems of residential buildings built in the 1950s and early 1960s in Vladimir were chosen as objects of research. Interpolation and extrapolation methods were used to build a predictive model of the residual life of a structure. Detailed calculations are given, which clearly show the possibility of using these methods. It is determined that the autoregression method (Burg method) shows good predictive results, correlating with experimental data from other studies and theoretical assumptions. Forecasting the remaining life of a structure is a key factor in ensuring the reliability and safety of buildings, as well as reducing future operating costs.
Architectural engineering. Structural engineering of buildings
This research aims to assess and quantify the significance of incorporating the seismic performance of global and local engineering demand parameters (EDPs) within the probabilistic frameworks, when structural pounding of adjacent buildings occurs. For this purpose, the seismic performance of 6-story and 12-story reinforced concrete (RC) RC frames subjected to floor-floor pounding is assessed. The pounding is caused by an adjacent shorter and stiffer structure with the top contact point at the middle of the tall building’s total height. Displacement-based and ductility-based EDPs are evaluated at different performance levels (PLs) and at different separation distances (dg). The seismic performance of the RC frames without considering pounding is also evaluated. Incremental dynamic analyses (IDAs) are performed, and probabilistic seismic demand models (PSDMs) are developed to establish fragility curves of the examined RC frames. The probability of earthquake-induced pounding between adjacent structures is properly involved with the median value of Sa,T1 that corresponds to an acceptable capacity level (acceptable PL) of an EDP. The results of this study indicate that excluding structural pounding consequences from the probabilistic frameworks related to the seismic risk of colliding buildings leads to unsafe seismic assessment or design provisions.
Leonhard Applis, Yuntong Zhang, Shanchao Liang
et al.
The growth of Large Language Model (LLM) technology has raised expectations for automated coding. However, software engineering is more than coding and is concerned with activities including maintenance and evolution of a project. In this context, the concept of LLM agents has gained traction, which utilize LLMs as reasoning engines to invoke external tools autonomously. But is an LLM agent the same as an AI software engineer? In this paper, we seek to understand this question by developing a Unified Software Engineering agent or USEagent. Unlike existing work which builds specialized agents for specific software tasks such as testing, debugging, and repair, our goal is to build a unified agent which can orchestrate and handle multiple capabilities. This gives the agent the promise of handling complex scenarios in software development such as fixing an incomplete patch, adding new features, or taking over code written by others. We envision USEagent as the first draft of a future AI Software Engineer which can be a team member in future software development teams involving both AI and humans. To evaluate the efficacy of USEagent, we build a Unified Software Engineering bench (USEbench) comprising of myriad tasks such as coding, testing, and patching. USEbench is a judicious mixture of tasks from existing benchmarks such as SWE-bench, SWT-bench, and REPOCOD. In an evaluation on USEbench consisting of 1,271 repository-level software engineering tasks, USEagent shows improved efficacy compared to existing general agents such as OpenHands CodeActAgent. There exist gaps in the capabilities of USEagent for certain coding tasks, which provides hints on further developing the AI Software Engineer of the future.
Muhammad Tayyab Khan, Zane Yong, Lequn Chen
et al.
Engineering drawings are fundamental to manufacturing communication, serving as the primary medium for conveying design intent, tolerances, and production details. However, interpreting complex multi-view drawings with dense annotations remains challenging using manual methods, generic optical character recognition (OCR) systems, or traditional deep learning approaches, due to varied layouts, orientations, and mixed symbolic-textual content. To address these challenges, this paper proposes a three-stage hybrid framework for the automated interpretation of 2D multi-view engineering drawings using modern detection and vision language models (VLMs). In the first stage, YOLOv11-det performs layout segmentation to localize key regions such as views, title blocks, and notes. The second stage uses YOLOv11-obb for orientation-aware, fine-grained detection of annotations, including measures, GD&T symbols, and surface roughness indicators. The third stage employs two Donut-based, OCR-free VLMs for semantic content parsing: the Alphabetical VLM extracts textual and categorical information from title blocks and notes, while the Numerical VLM interprets quantitative data such as measures, GD&T frames, and surface roughness. Two specialized datasets were developed to ensure robustness and generalization: 1,000 drawings for layout detection and 1,406 for annotation-level training. The Alphabetical VLM achieved an overall F1 score of 0.672, while the Numerical VLM reached 0.963, demonstrating strong performance in textual and quantitative interpretation, respectively. The unified JSON output enables seamless integration with CAD and manufacturing databases, providing a scalable solution for intelligent engineering drawing analysis.
Model-driven engineering (MDE) is believed to have a significant impact in software quality. However, researchers and practitioners may have a hard time locating consolidated evidence on this impact, as the available information is scattered in several different publications. Our goal is to aggregate consolidated findings on quality in MDE, facilitating the work of researchers and practitioners in learning about the coverage and main findings of existing work as well as identifying relatively unexplored niches of research that need further attention. We performed a tertiary study on quality in MDE, in order to gain a better understanding of its most prominent findings and existing challenges, as reported in the literature. We identified 22 systematic literature reviews and mapping studies and the most relevant quality attributes addressed by each of those studies, in the context of MDE. Maintainability is clearly the most often studied and reported quality attribute impacted by MDE. Eighty out of 83 research questions in the selected secondary studies have a structure that is more often associated with mapping existing research than with answering more concrete research questions (e.g., comparing two alternative MDE approaches with respect to their impact on a specific quality attribute). We briefly outline the main contributions of each of the selected literature reviews. In the collected studies, we observed a broad coverage of software product quality, although frequently accompanied by notes on how much more empirical research is needed to further validate existing claims. Relatively, little attention seems to be devoted to the impact of MDE on the quality in use of products developed using MDE.
Haoran Liang, Yufa Zhou, Mohammad Talebi Kalaleh
et al.
We introduce $\textbf{MASSE}$, the first Multi-Agent System for Structural Engineering, effectively integrating large language model (LLM)-based agents with real-world engineering workflows. Structural engineering is a fundamental yet traditionally stagnant domain, with core workflows remaining largely unchanged for decades despite its substantial economic impact and global market size. Recent advancements in LLMs have significantly enhanced their ability to perform complex reasoning, long-horizon planning, and precise tool utilization -- capabilities well aligned with structural engineering tasks such as interpreting design codes, executing load calculations, and verifying structural capacities. We present a proof-of-concept showing that most real-world structural engineering workflows can be fully automated through a training-free LLM-based multi-agent system. MASSE enables immediate deployment in professional environments, and our comprehensive validation on real-world case studies demonstrates that it can reduce expert workload from approximately two hours to mere minutes, while enhancing both reliability and accuracy in practical engineering scenarios.
This paper introduces Design for Sensing and Digitalisation (DSD), a new engineering design paradigm that integrates sensor technology for digitisation and digitalisation from the earliest stages of the design process. Unlike traditional methodologies that treat sensing as an afterthought, DSD emphasises sensor integration, signal path optimisation, and real-time data utilisation as core design principles. The paper outlines DSD's key principles, discusses its role in enabling digital twin technology, and argues for its importance in modern engineering education. By adopting DSD, engineers can create more intelligent and adaptable systems that leverage real-time data for continuous design iteration, operational optimisation and data-driven predictive maintenance.
Thiago Barradas, Aline Paes, Vânia de Oliveira Neves
The effective execution of tests for REST APIs remains a considerable challenge for development teams, driven by the inherent complexity of distributed systems, the multitude of possible scenarios, and the limited time available for test design. Exhaustive testing of all input combinations is impractical, often resulting in undetected failures, high manual effort, and limited test coverage. To address these issues, we introduce RestTSLLM, an approach that uses Test Specification Language (TSL) in conjunction with Large Language Models (LLMs) to automate the generation of test cases for REST APIs. The approach targets two core challenges: the creation of test scenarios and the definition of appropriate input data. The proposed solution integrates prompt engineering techniques with an automated pipeline to evaluate various LLMs on their ability to generate tests from OpenAPI specifications. The evaluation focused on metrics such as success rate, test coverage, and mutation score, enabling a systematic comparison of model performance. The results indicate that the best-performing LLMs - Claude 3.5 Sonnet (Anthropic), Deepseek R1 (Deepseek), Qwen 2.5 32b (Alibaba), and Sabia 3 (Maritaca) - consistently produced robust and contextually coherent REST API tests. Among them, Claude 3.5 Sonnet outperformed all other models across every metric, emerging in this study as the most suitable model for this task. These findings highlight the potential of LLMs to automate the generation of tests based on API specifications.
Armin Ariamajd, Raquel López-Ríos de Castro, Andrea Volkamer
The increasing importance of Computational Science and Engineering has highlighted the need for high-quality scientific software. However, research software development is often hindered by limited funding, time, staffing, and technical resources. To address these challenges, we introduce PyPackIT, a cloud-based automation tool designed to streamline research software engineering in accordance with FAIR (Findable, Accessible, Interoperable, and Reusable) and Open Science principles. PyPackIT is a user-friendly, ready-to-use software that enables scientists to focus on the scientific aspects of their projects while automating repetitive tasks and enforcing best practices throughout the software development life cycle. Using modern Continuous software engineering and DevOps methodologies, PyPackIT offers a robust project infrastructure including a build-ready Python package skeleton, a fully operational documentation and test suite, and a control center for dynamic project management and customization. PyPackIT integrates seamlessly with GitHub's version control system, issue tracker, and pull-based model to establish a fully-automated software development workflow. Exploiting GitHub Actions, PyPackIT provides a cloud-native Agile development environment using containerization, Configuration-as-Code, and Continuous Integration, Deployment, Testing, Refactoring, and Maintenance pipelines. PyPackIT is an open-source software suite that seamlessly integrates with both new and existing projects via a public GitHub repository template at https://github.com/repodynamics/pypackit.
Successfully engineering interactive industrial DTs is a complex task, especially when implementing services beyond passive monitoring. We present here an experience report on engineering a safety-critical digital twin (DT) for beer fermentation monitoring, which provides continual sampling and reduces manual sampling time by 91%. We document our systematic methodology and practical solutions for implementing bidirectional DTs in industrial environments. This includes our three-phase engineering approach that transforms a passive monitoring system into an interactive Type 2 DT with real-time control capabilities for pressurized systems operating at seven bar. We contribute details of multi-layered safety protocols, hardware-software integration strategies across Arduino controllers and Unity visualization, and real-time synchronization solutions. We document specific engineering challenges and solutions spanning interdisciplinary integration, demonstrating how our use of the constellation reporting framework facilitates cross-domain collaboration. Key findings include the critical importance of safety-first design, simulation-driven development, and progressive implementation strategies. Our work thus provides actionable guidance for practitioners developing DTs requiring bidirectional control in safety-critical applications.
Large language models exhibit intelligence without genuine epistemic understanding, exposing a key gap: the absence of epistemic architecture. This paper introduces the Structured Cognitive Loop (SCL) as an executable epistemological framework for emergent intelligence. Unlike traditional AI research asking "what is intelligence?" (ontological), SCL asks "under what conditions does cognition emerge?" (epistemological). Grounded in philosophy of mind and cognitive phenomenology, SCL bridges conceptual philosophy and implementable cognition. Drawing on process philosophy, enactive cognition, and extended mind theory, we define intelligence not as a property but as a performed process -- a continuous loop of judgment, memory, control, action, and regulation. SCL makes three contributions. First, it operationalizes philosophical insights into computationally interpretable structures, enabling "executable epistemology" -- philosophy as structural experiment. Second, it shows that functional separation within cognitive architecture yields more coherent and interpretable behavior than monolithic prompt based systems, supported by agent evaluations. Third, it redefines intelligence: not representational accuracy but the capacity to reconstruct its own epistemic state through intentional understanding. This framework impacts philosophy of mind, epistemology, and AI. For philosophy, it allows theories of cognition to be enacted and tested. For AI, it grounds behavior in epistemic structure rather than statistical regularity. For epistemology, it frames knowledge not as truth possession but as continuous reconstruction within a phenomenologically coherent loop. We situate SCL within debates on cognitive phenomenology, emergence, normativity, and intentionality, arguing that real progress requires not larger models but architectures that realize cognitive principles structurally.
The demand for transparent building envelopes, particularly glass facades, is rising in modern architecture. These facades are expected to meet multiple objectives, including aesthetic appeal, durability, quick installation, transparency, and both economic and ecological efficiency. At the heart of facade design, particularly for structural glass elements, lies the assurance of structural integrity for ultimate and serviceability limit states with a requisite level of reliability. However, current structural engineering assessments for glass and glass laminate designs, especially in the geometrically non-linear setting, are time-consuming and require significant expertise. This study develops a customized Mixture-of-Experts (MoE) neural network architecture to overcome current limitations. It calibrates it on synthetically generated stress and deformation data obtained via parametrized Finite-Element-Analysis (FEA) of glass and glass laminate structures under both geometrically linear and nonlinear conditions for several joint support and loading conditions. Our findings reveal that the MoE model outperforms baseline models in predicting laminate deflections and stresses, offering a substantial increase in computational efficiency, compared to traditional linear and non-linear FEA, at high accuracy. The MoE is integrated within a novel web-based glass design and verification tool called Strength Lab AI and provided to the engineering public for future use. These results have profound implications for advancing engineering practice, offering a robust tool for the intricate structural design and analysis of glass and glass laminate structures.
Many buildings during their operational period incur damage of different origin: man-made, natural, operational, etc. Dynamic tests are performed for detailed assessment of the technical condition of buildings and structures in accordance with the regulatory documents for general analysis of the building damage state. In a large number of papers, the results of comparison of full-scale tests and numerical analysis using finite element method are presented. When analyzing the results, it can be concluded that the dynamic method is reliable, but has several limitations. The advantage of the dynamic method of building damage assessment is the possibility to adjust finite element models in software systems taking into account results obtained from in-situ tests, which allows to obtain more accurate results for the assessment of bearing capacity under seismic loading. To examine the effect of damage to buildings on their seismic resistance, an experiment with corrosiondamaged reinforced concrete columns was performed. The result of the first stage of the experiment is the assessment of the change in dynamic characteristics (eigenfrequency, vibration decrement, vibration damping coefficient, etc.) of reinforced concrete column specimens subjected to corrosion damage.
Architectural engineering. Structural engineering of buildings
Sadiq Sadiq Mussadaq M Baqer, Hatem A Gzar, Qasim M Jani
et al.
In this study, we investigated Terasil blue dye absorption on modified rice husk through batch and continuous trials. In continuous mode includes experimental tests in an inverse fluidized bed teqnique at various times and under various operating conditions (bed height, initial concentration, and varying flow rate) were investigated. The effect of various factors like pH, contact time, agitation speed and particle size on the removal efficiency (%) the Terasil blue dye were thoroughly investigated. The maximum removal efficiency (%) was achieved up to pH 7.0. Removal efficiency (%) increased with increasing contact time. The maximum removal efficiency (%) was achieved @200 RPM (rate per minute). Increasing in the particle size caused decreased in the removal efficiency (%). In batch experiments the Freundlich, and Temkin models showed good agreement with R2 value while the Langmuir model had moderate agreement. The value of qe is 1.73 mg/g under specific conditions; the Langmuir model provides q-max of 0.0078 mg/g and K-L of 0.0801 L/mg.Ongoing tests conducted in an inverse fluidized bed offer valuable insights into the hydrodynamic behavior of the system.
The collected data effectively demonstrates the variations in pressure drops and bed heights. There is a positive correlation between bed height and fluid velocity, suggesting a significant association within the dynamics of fluidized beds.
Piotr Sowinski, Ignacio Lacalle, Rafael Vano
et al.
The landscape of computing technologies is changing rapidly, straining existing software engineering practices and tools. The growing need to produce and maintain increasingly complex multi-architecture applications makes it crucial to effectively accelerate and automate software engineering processes. At the same time, artificial intelligence (AI) tools are expected to work hand-in-hand with human developers. Therefore, it becomes critical to model the software accurately, so that the AI and humans can share a common understanding of the problem. In this contribution, firstly, an in-depth overview of these interconnected challenges faced by modern software engineering is presented. Secondly, to tackle them, a novel architecture based on the emerging WebAssembly technology and the latest advancements in neuro-symbolic AI, autonomy, and knowledge graphs is proposed. The presented system architecture is based on the concept of dynamic, knowledge graph-based WebAssembly Twins, which model the software throughout all stages of its lifecycle. The resulting systems are to possess advanced autonomous capabilities, with full transparency and controllability by the end user. The concept takes a leap beyond the current software engineering approaches, addressing some of the most urgent issues in the field. Finally, the efforts towards realizing the proposed approach as well as future research directions are summarized.
L. Sanhudo, Nuno M. M. Ramos, João Poças Martins
et al.
Abstract Building Information Modeling (BIM), as a rising technology in the Architecture, Engineering and Construction (AEC) industry, has been applied to various research topics from project planning, structural design, facility management, among others. Furthermore, with the increasing demand for energy efficiency, the AEC industry requires an expeditious energy retrofit of the existing building stock to successfully achieve the 2020 Energy Strategy targets. As such, this article seeks to survey the recent developments in the energy efficiency of buildings, combining energy retrofitting and the technological capabilities of BIM, providing a critical exposition in both engineering and energy domains. The result is a thorough review of the work done by other authors in relevant fields, comprising the entire spectrum from on-site data acquisition, through the generation of Building Energy Models (BEM), data transfer to energy analysis software and, finally, the identification of major issues throughout this process. Additionally, a BIM-based methodology centered on the acquired knowledge is presented. Solutions for as-built data acquisition such as laser scanning and infrared thermography, and on-site energy tests that benefit the acquisition of energy-related data are explored. The most predominant BIM software regarding not only energy analysis but also model development is examined. In addition, interoperability restrictions between BIM and energy analysis software are addressed using the Industry Foundation Classes (IFC) and Green Building Extensible Markup Language (gbXML) schemes. Lastly, the article argues the future innovations in this subject, predicting future trends and challenges for the industry.