The International Association for Automation and Robotics in Construction
Search papers
Abstract: The International Association for Automation and Robotics in Construction (IAARC) and the Local Organizing Committee are pleased to present the Proceedings of the 43rd International Symposium on Automation and Robotics in Construction (ISARC 2026). This year's symposium draws inspiration from the vibrant spirit, innovation, and cultural diversity of its host ...
Abstract: On-site sandwich-panel assembly demands safe, precise manipulation by heavy?duty equipment operating under dynamic, partially known constraints. We present motion-control design and validation for a retrofitted rotating telehandler. A ROS2-compatible Unity digital twin is used for safe, repeatable tuning. The twin embeds joint-level non-linearities (backlash, friction, compliance, saturation) to reduce sim-to-real ...
Abstract: Enabling robust loco-manipulation for transporting unwieldy objects is a critical prerequisite for deploying humanoid robots in construction. However, this task is complicated by severe physical constraints, including irregular object geometries, limited actuator torque, and high moments of inertia caused by eccentric centers of mass. These factors induce complex robot-object dynamics ...
Abstract: This research proposes a plug-and-play semantic execution framework that enables Digital Twin--oriented automation to be configured through reusable domain knowledge rather than project-specific code. The framework is built on Domain Packs?modular semantic units encapsulating ontologies, data mappings, reasoning rules, workflow templates, and provenance policies?and a set of domain-agnostic agents operating ...
Abstract: Digital twins (DTs) have become a cornerstone of cyberphysical systems, enabling real-time monitoring, control, and predictive analytics through continuous data exchange between physical assets and their virtual counterparts. While extensively applied in manufacturing, their adoption in off-site wood construction remains limited, particularly with respect to ...
Abstract: With recent advancements in robotics, robots have been successfully utilized to automate processes in various sectors. However, their use in the construction industry has not mirrored this trend, due to the complexity of construction processes and the fact that they take place on highly dynamic construction sites. Therefore, construction robots ...
Abstract: Understanding on-site activities is fundamental for supporting construction planning, coordination, and performance assessment; however, this task is commonly performed through manual interpretation by human supervisors, limiting information integration with autonomous systems. Scene graph generation (SGG) provides a structured representation of visual scenes, but existing image-based methods rely heavily on supervised ...
Abstract: As infrastructure systems such as bridges continue to age, maintenance engineers are increasingly required to make timely and consistent decisions based on large volumes of unstructured maintenance documents. In current practice, decision-making relies heavily on inspectors' experience and manual interpretation of guidelines, inspection reports, and past cases, often resulting in ...
Abstract: Precast bridge construction accelerates the construction process and minimizes traffic disruptions. However, the onsite installation demands extensive planning efforts from experienced crane operators to ensure safe and efficient assembly operation. This study presents an integrated IFC-based simulation framework that combines automated BIM information extraction with heuristic crane planning algorithms to ...
Abstract: Construction safety inspection remains a critical yet challenging task due to complex site environments, dynamic worker activities, and frequent visual occlusions. While recent advances in computer vision and vision-language models (VLMs) have enabled automated safety reasoning, most existing systems rely on single-agent, single-view observations, which are inherently vulnerable to occlusion ...
Abstract: Structural Support Assemblies (SSAs) connect Mechanical, Electrical, and Plumbing (MEP) services to building structures under strict geometric, catalog-defined compatibility, and structural constraints. Despite their importance, SSA layouts and component selections are still largely determined through manual engineering judgment, even in Building Information Modeling-enabled workflows. Existing digital tools primarily verify or ...
Abstract: //To address limitations of traditional manual patrols on construction sites, such as inconsistent and inaccurate hazard identification, unmanned mobile robots can be deployed. However, specific behavioral parameters of unmanned mobile robots may distract workers' attention. Although related research has been conducted, it remains a critical knowledge gap which robot dogs' ...
Abstract: Construction workflows are restricted by multiple constraints, factors that limit when and how tasks can be performed. Among them, spatial constraints are particularly important as workers and equipment often compete for limited areas, and insufficient clearance can make tasks unsafe or reduce productivity. However, existing planning tools, such as the ...
Abstract: Sudden incidents can lead to severe consequences, and traditional emergency-response systems often suffer from incomplete information and delayed decision-making. To address this challenge, we propose an intelligent emergency response system that integrates structured knowledge modeling with large language models (LLMs). The system constructs a knowledge base comprising experiential emergency guidelines ...
Abstract: Lean principles have long supported digital transformation in manufacturing. The architecture, engineering, construction and operations (AECO) sector has increasingly adopted them for off-site fabrication. Yet, AECO projects usually involve bespoke designs, unique site conditions and project-specific processes; and the role of lean in advancing emerging robotic fabrication remains under-investigated. ...
Abstract: In this paper, we propose a modeling approach for feedforward control of vehicle velocity in an autonomous bulldozer. Unlike tracked vehicles in which the left and right tracks can be commanded independently, a bulldozer is operated by two inputs?acceleration/deceleration and steering?analogous to throttle and steering-wheel inputs in passenger cars. Accordingly, ...
Abstract: Robotic automation of rebar cage fabrication continues to face challenges due to its high spatial complexity, as dense reinforcement layouts, narrow clearances and irregular geometry restrict robot maneuverability. The groundwork for intelligent, perception-aware, and adaptable construction automation framework is missing in the literature. This research explores how the complexity of ...
Abstract: Mobile 3D printing on unstructured terrain remains challenging due to the conflict between platform mobility and deposition precision. Existing gantry-based systems achieve high accuracy but lack mobility, while mobile platforms struggle to maintain print quality on uneven ground. We present a framework that tightly integrates AI-driven disturbance prediction with multi-modal ...
Abstract: Manual assembly in construction and manufacturing currently relies on static instructions (manuals, SOPs) that are brittle to variability and lack real-time feedback. To address this, we present a laser-based multimodal co-pilot that delivers dynamic, closed-loop guidance in non-wearable settings. The proposed framework compiles heterogeneous documentation (manuals, CAD/BIM) into a unified ...
Abstract: Robotic automation in construction remains limited by the complexity of physically interactive tasks and the variability of real job sites. This study presents a hardware-validated proof-of-concept for imitation learning-based Physical AI on a small-scale excavation task. A modified SO?101 robotic arm was equipped with a 3D-printed excavation bucket, and 50 ...
Abstract: Mobile robots deployed in cluttered environments like in construction must navigate safely in cluttered spaces while satisfying state, input, and collision-avoidance constraints. Global path planners can generate optimal paths but do not guarantee safe execution, whereas constrained control methods ensure safety but often suffer from local minima. This paper proposes ...
Abstract: Resilience to unexpected failures is critical for multi-agent collaboration in dynamic construction environments. Existing recovery strategies often rely on rigid contingency rules or computationally expensive replanning, which struggle with heterogeneous fleets and complex dependencies. We propose an adaptive recovery framework leveraging Large Language Models (LLMs) to generate executable recovery plans ...
Abstract: Construction robotics has advanced rapidly in recent years, yet the deployment of autonomous and multi?robot systems at scale remains constrained by computational challenges arising from complex, dynamic, and uncertain construction environments. Core problems such as task allocation, construction sequencing, motion planning, and multi?robot coordination often exhibit combinatorial complexity, limiting the ...
Abstract: Construction tasks expose workers to fluctuating cognitive demands that influence attention, accuracy, and fatigue. Understanding cognitive performance requires examining not only overload triggers, but also mechanisms that stabilize performance and support recovery. This paper explores the positive side of cognitive regulation which includes the task characteristics, and situational/behavioral factors that ...
Abstract: Multi-robot construction tasks remain difficult to automate using pre-programmed control due to environmental variability, tight tolerances, and the need for continuous coordination during execution. This paper presents preliminary work toward adaptive dual-robot façade assembly, using anchor-bracket installation as a representative case. Through simulation-based studies, we validate the feasibility of pre-determined ...
Abstract: Task planning is essential for the operation of au- tonomous machines. For autonomous construction vehi- cles, such as drilling excavators, these tasks not only need to be performed in an efficient order but also must take the terrain into account. This work presents an approach ...
Abstract: On-site inspection of mechanical, electrical, and plumbing (MEP) installations is often inefficient because 2D drawings are unintuitive, making it difficult for engineers to interpret complex spatial relationships and retrieve relevant information quickly. To bring BIM information to the point of work, this study proposed a BIM-integrated augmented reality (AR) multi-modal ...
Abstract: Expectations are growing for low-orbit satellite communication network to enable remote operation of construction machinery in mountainous areas and disaster recovery sites where public communication infrastructure is difficult to establish. However, transmitting multiple camera feeds for remote operation of multiple machines poses challenges, as bandwidth constraints and packet loss make ...
Abstract: Tower cranes are underactuated, nonlinear systems in which rapid trolley and slew motions inevitably cause payload swing. An effective controller is crucial to ensure operational safety and efficiency, as uncontrolled payload swing can lead to hazardous conditions and reduced productivity. This paper addresses the problem of rapid and restricted swing ...
Abstract: Accurate and machine-readable indoor maps are essential for autonomous systems, such as mobile robots, to operate effectively. Existing traditional mapping methods, such as manual piloting or Simultaneous Localization and Mapping (SLAM), are often time-consuming, lack semantic details, and are affected by environmental changes. Building Information Modeling (BIM) has recently been ...
Abstract: Robotic navigation in construction environments remains challenging due to unstructured layouts, frequent environmental changes, and weak semantic regularities, which limit the applicability of current navigation approaches. Recent vision-language models offer new opportunities for semantic goal-oriented navigation by enabling robots to reason over high-level language instructions and visual observations. However, their ...
Abstract: This paper proposes a localization-based interface for semi-automated crane movement, allowing the crane to navigate to the operator's position via a low-cost Ultra-Wideband (UWB) tag. This could reduce the cognitive load induced by traditional interfaces and improve safety and productivity. Due to the hardware-specific challenges to access Channel Impulse Response ...
Abstract: The U.S. construction industry continues to face persistent labor shortages that contribute to project delays and productivity loss. Mobile robots are a promising way to supplement human labor by automating repetitive and physically demanding tasks. Among these platforms, quadruped robots are well suited for construction sites because legged locomotion enables ...
Abstract: Despite the existence of modern capable robots, crews still carry plywood by hand and use nail guns to fasten it on a typical timber-frame job. Panel handling and fastening are obvious candidates for automation because they involve repetitive, labor-demanding tasks that put workers in danger. However, in current practice, the ...
Abstract: Off-site light-gauge steel LGS frame manufacturing plays a key role in modern panelized construction due to its speed, precision, and reduced on-site labor. How- ever, screw fastening defects remain a persistent issue that affects structural integrity, quality compliance, and pro- duction efficiency. This paper presents ...
Abstract: Prefabricated buildings are increasingly adopted for their efficiency and environmental benefits. However, during hoisting and installation, accurate docking between bottom sleeves and pre-reserved steel bar regions is often hindered by limited visibility and poor controllability of manual fine-tuning, reducing installation quality and efficiency. This paper proposes a four-cable-driven parallel robot ...
Abstract: The deployment of fully autonomous construction equipment for earthworks has gained significant interest in both academia and industry. This is driven by rapid advancements in artificial intelligence skilled labour shortages and the need to drive efficiency, safety and productivity. Despite successful trials, questions remain about their wider impact on the ...
Abstract: Efficient fleet scheduling in urban infrastructure maintenance and inspection services requires a delicate balance between task productivity and operational safety. However, traditional centralized algorithms often struggle with the stochastic nature of dynamic urban environments, leading to communication latency and collision risks. To bridge the gap between global platform planning and ...
Abstract: The construction industry has one of the highest rates of workplace fatalities and severe injuries. Effective, high-fidelity safety training is not merely a regulatory requirement but a moral imperative. While Extended Reality (XR) has proven to improve the training results in construction safety training, a bottleneck is its resource-intensive content ...
Abstract: Structural inspections increasingly rely on unmanned aerial vehicles (UAVs) to improve access and safety, however, UAV-based bridge inspections often suffer from inconsistent data coverage, pilot-dependent flight paths, and limited repeatability across inspection campaigns. This study presents a fully autonomous drone inspection framework that integrates 3D Gaussian Splatting (3DGS) with optimization-based ...
Abstract: This work targets low mechanization, safety risks, and labor-intensive foundation construction for UHV transmission lines over complex terrains. We develop a wheel-leg hybrid platform integrating a multi-DOF chassis, modular drilling system, and intelligent electro-hydraulic control for autonomous terrain adaptation, real-time posture adjustment, precise leveling, and remote operation. A full system ...
Abstract: This paper introduces an orchestration framework for the Information Backbone for Robotic Construction (IBRC), a unified platform that connects fragmented planning and on-site robotic construction workflows. Construction sites require continuous coordination among heterogeneous robots, humans, machines, and digital models, yet existing solutions typically address isolated integrations (e.g., BIM-to-robot pipelines or ...
Abstract: Assembly tasks in constrction are characterized by customized components, complex connection relationships, and dynamic environments, making conventional approaches designed for fixed workflows difficult to reliably reuse. This paper proposes an assembly agent framework that integrates Rhino spatial data with multimodal large language models, enabling end-to-end closed-loop control from digital models ...
Abstract: Mobile robots are a promising platform for large-scale on-site construction, but unlike fixed-base robots, they lack mechanical reference links between workspaces. On evolving construction sites, this creates a fundamental challenge: maintaining consistent alignment between digital planning models and physical execution across multiple robot positions and operational scales. This paper presents ...
Abstract: Abstract - This paper presents a collaborative manipulation frame- work for automated stacking of large masonry blocks on-site. Our approach utilizes an automated hoist for lifting the blocks and transporting them to the deposition site, combined with a ground-based manipulator for catching and precisely guiding ...
Abstract: This contribution presents the approach for the digital quality assessment of surfaces of additively manufactured concrete components. It is based on process-integrated sensor technology, surface-structural quality criteria, and integration into the cybernetic control loop of the printing process. A continuous target-actual comparison of the constructed geometry enables robotically adapted "first-time-right" ...
Abstract: Automating tolerance-critical construction assembly, particularlyprefabricated window installation, remains a persistent challenge due to millimeter-scale clearances, complex multi-surface friction, and the risks of hardware damage. In the final insertion stage, small pose errors rapidly escalate into jamming states that traditional open-loop planning struggles to resolve. While Reinforcement Learning (RL) ...
Abstract: This paper presents a scaled mobile robotic additive manufacturing (AM) system developed to investigate system-level feasibility toward underwater applications, with a specific focus on concrete-based material deposition. The proposed platform integrates a tracked mobile base, a lightweight robotic manipulator with offline-programmed motion, and a syringe-based extrusion system designed for cementitious ...
Abstract: The need to handle heavy loads safely and precisely in complex and unstructured environments is driving the increasing use of robotic systems in construction and field robotics. Curtain wall panel installation is one representative task that can benefit from such systems, requiring the accurate placement of large and heavy façade ...
Abstract: Automatic rebar tying using a vision-guided robotic system requires a reliable perception of rebar intersections within densely arranged rebar meshes. However, major challenges include the scarcity of large-scale, precisely annotated visual datasets for training perception models and the limited evaluation of models trained on synthetic data in the context of ...
Abstract: As the construction industry undergoes a digital transformation, there is an increasing demand to transition from rigid manual fabrication to intelligent, flexible multi-robotic systems capable of managing complex design constraints. This research presents a generalized framework for a multi-robotic cell designed for Light Gauge Steel (LGS) assembly. The methodology utilizes ...
Abstract: Mobile legged robots offer unique advantages for on-site construction, including the ability to traverse unstructured terrain, climb stairs, and operate independently between spatially separated work zones. This paper presents a proof-of-concept system for autonomous brick stacking using a Boston Dynamics Spot quadrupedal robot equipped with a manipulator arm. The system ...
Abstract: The on-site assembly of prefabricated modules relies heavily on manual labor. However, it's inefficient and exposes workers to significant safety risks, remaining a major bottleneck for modular construction (MC) assembly. This paper proposes a cooperative motor-thruster vertical orientation control system (CM-TVOCS) for suspended modules during lifting and positioning. A rotary ...
Abstract: Autonomous rebar tying requires precise mobile base docking relative to the rebar cage to ensure effective manipulation. However, this process is challenged by the sparse, repetitive, and texture-less nature of rebar structures, which leads to perceptual aliasing and measurement instability in conventional LiDAR and visual perception systems. To address this ...
Abstract: Alignment control of grouted-sleeve connections is a critical yet labor-intensive step in precast concrete (PC) assembly. Current practice relies heavily on manual measurements before lifting and on experience-based corrections during hoisting, which makes deviation quantification slow, subjective, and difficult to digitalize. This paper presents a point cloud-image fusion method for ...
Abstract: Reliable on-site positioning is essential for robotic installation of autoclaved lightweight concrete (ALC) wall panels. A significant challenge arises from the use of close-range wide-angle binocular vision, which frequently produces marker images suffering from low spatial resolution and low contrast. These degraded marker images increase positioning errors, which propagate through ...
Abstract: Humanoid robots offer a promising solution for construction and building operation tasks because they can operate in environments designed for humans without requiring major changes to infrastructure or workflows. However, deploying humanoids in real construction and building environments remains challenging due to safety risks, complex interactions with the built environment, ...
Abstract: Traditional port inspections are highly dependent on manual labor and static sensors, leading to high operational costs and fragmented data. These limitations hinder effective data integration and advanced analysis, creating a critical need for an automated inspection framework to improve safety and efficiency. This paper proposes an automated safety inspection ...
Abstract: As the construction industry transitions toward Construction 4.0, there is an increasing demand for automated technologies to replace labour-intensive manual coating processes at construction sites. Traditional spraying systems are often limited to single-mode operation and manual control, making them insufficient for the automated application of diverse material properties found on ...
Abstract: The increasing frequency of large-scale public events has heightened the risk of crowd crush incidents, which present significant public safety challenges. Despite advances in research on crowd behavior, existing methods, including purely virtual simulations, bear fundamental limitations in capturing the critical physical dynamics?such as contact forces and inertia?that govern high-density ...
Abstract: Robust perception of rebar joints is challenging due to geometric variability, irregular intersection topologies, and real-world sensing imperfections. Existing methods largely rely on appearance-driven recognition and topology-specific supervision, which limits generalization. We reformulate rebar joint perception as a geometry-dominated structural understanding problem and propose a two-stage learning pipeline that biases ...
Abstract: Rebar tying operations are highly repetitive and labor-intensive, presenting significant potential for automation. Although several rebar tying robots have already been deployed, in scenarios involving vertical or inclined rebar cages, conventional single-degree-of-freedom actuators prove inadequate, thereby necessitating the adoption of multi-axis robotic arm solutions. This introduces the novel problem of ...
Abstract: Unmanned Aerial Vehicles (UAV) images can streamline reinforced-concrete inspections, yet reliable automatic rebar counting is still challenging because rebars often occupy only a few pixels, appear in dense groups, and are frequently obscured or visually blended with background textures, shadows, and occlusions. To address this small-object setting, we develop an ...
Abstract: Robotic layout projection has been increasingly adopted to support construction layout tasks while reducing manual effort. However, existing solutions still require substantial human intervention and are largely limited to single, horizontal surfaces. To address these limitations, this paper proposes a BIM-driven robotic framework for autonomous construction layout projection. The framework ...
Abstract: Inspection and maintenance of confined spaces in critical infrastructure pose major challenges for the Architecture, Engineering, Construction, and Operation (AECO) industry, including high safety risks, substantial inefficiencies, and heavy dependence on manual labor. Despite the promise of Industry 4.0 technologies, the adoption of autonomous systems remains limited due to insufficient ...
Abstract: Construction robotics requires effective coordination strategies for heterogeneous robot teams performing long-horizon and complex tasks. Traditional approaches using Dynamic Programming and Reinforcement Learning (RL) have dominated multi-robot task allocation, yet emerging Large Language Model (LLM) solutions demonstrate a more productive allocation scheme. To further improve LLM-based multi-robot task allocation for ...
Abstract: Water ponding or puddle is one of the key factors contributing to poor housekeeping conditions on construction sites, which can significantly hinder safety, operational efficiency, and creating favourable conditions for mosquito breeding.[YG1.1] Traditional housekeeping monitoring methods, which rely on manual observations, are often ineffective. Moreover, although computer vision has shown ...
Abstract: Robotic fabrication holds significant potential for improving productivity in construction, yet its adoption remains limited by the task variability, high-mix low-volume production, and the tight coupling requirement between design and production. Furthermore, translating design intent into physical production often relies on fragmented software ecosystems operated by different specialists, increasing coordination ...
Abstract: Unmanned aerial vehicles (UAVs) are increasingly used in construction for inspection, monitoring, and safety management. While these applications offer operational benefits, recent studies indicate that drone presence and motion can distract on-site workers and impair situational awareness. Existing construction drone path planning methods, however, primarily emphasize efficiency, coverage, and collision ...
Abstract: Robotic assembly of timber structures promises increased efficiency and sustainability in construction. However, conventional timber joints often cause high assembly difficulty for robots (e.g. bolts) and use non-reversible connectors (e.g. nails or glue). This paper presents a novel curved mortise-tenon timber joint designed for robotic non-sequential assembly and a dual-arm ...
Abstract: The construction industry is facing a severe labor shortage, and robotic automation systems are increasingly being deployed to support construction tasks. However, most existing robotic navigation systems in construction environments rely on manually designed modular pipelines, where each component is developed independently. Errors introduced by earlier modules are difficult to ...
Abstract: Masonry construction, the assembly of discrete units into load-bearing structures, offers distinctive advantages for sustainable building: local material sourcing, adaptability to diverse resources, and reduced embodied carbon when using minimally processed materials. This review traces the evolution of robotic masonry from early brick-laying machines of the 1980s to emerging systems ...
Abstract: Concrete 3D printing has significant potential to automate construction by improving productivity and reducing project costs. However, auxiliary workflows such as filler placement, reinforcement positioning, and grouting are typically performed manually or through semi-mechanized processes, limiting overall automation benefits. This study proposes an automated methodology for the fabrication of reinforced ...
Abstract: Developing autonomous construction robots is crucial for alleviating the compounding pressures of housing supply constraints and persistent labor shortages. Imitation learning has been increasingly adopted in the development of autonomous construction robots in recent years. However, the efficient collection of high-quality expert demonstrations from which robots learn remains a key ...
Abstract: Construction sites remain among the least productive and most hazardous work environments. AI-driven autonomous and semi-autonomous machinery could improve productivity and safety, yet deployment is constrained by reliability and assurance under dynamic operational design domains (ODDs). This paper presents an evidence-weighted scoping review of on-site AI-enabled autonomy and robotics, reported ...
Abstract: In recent years, the Japanese construction industry has been facing a serious labor shortage, and the practical realization of automated construction has become increasingly necessary. However, except for dam construction sites, automated earthwork has not yet been realized in general earthwork projects. To address this issue, this study ...
Abstract: Modular construction (MC) has gained increasing adoption due to its advantages in productivity, standardization, and off-site fabrication. However, the on-site installation of MC modules remains labor-intensive and highly dependent on repetitive manual leveling using a hand chain. This manual process introduces operational risks and prolongs installation time. To address these ...
Abstract: Bricklaying faces significant labor shortages, for which Mobile Construction Robots (MCRs) offer a promising solution. However, the absence of many building elements and hazards in brickwork environments, such as floor and exterior wall openings, which may be undetectable to robot sensors, poses safety challenges to the navigation system. This paper ...
Abstract: Quadruped robots are used for primary searches during the early stages of indoor fires. A typical primary search involves quickly and thoroughly looking for victims under hazardous conditions and monitoring flammable materials. However, situational awareness in complex indoor environments and rapid stair climbing and descending across different staircases remain the ...
Abstract: Human-robot collaboration in construction is often challenged by limited robot-to-human communication and the need to adapt to tolerance accumulation arising from material and assembly uncertainties. We present an adaptive human-robot collaborative workflow for masonry construction that addresses communication limitations and tolerance accumulation, demonstrated through a brickwork case study in which ...
Abstract: In prefabricated construction sites, tower cranes are required to operate safely in confined workspaces where site layouts and surrounding activities change over time. Reliable trajectory planning is therefore essential for maintaining safety margins, ensuring smooth motion, and improving operational efficiency. This paper proposes a configuration-space-based genetic optimization framework for tower ...
Abstract: The shift towards prefabricated and modular construction has increased reliance on tower cranes for handling heavier loads, making operational precision more critical than ever. However, traditional path planning methods often lack the adaptability and real-time performance required to prevent accidents in dynamic construction environments. This study develops a reinforcement learning ...
Abstract: Additive construction, also known as 3D concrete printing (3DCP), has emerged as a promising construction method with the potential to address industry challenges, such as low productivity and a shortage of skilled labor. In 3DCP, it is critical to have a reliable, real-time inspection process that continuously monitors material deposition, ...
Abstract: Automated inspection of indoor construction sites is pivotal for enhancing project quality management and operational safety. Mobile inspection robots possess the potential to gradually replace human workers. However, characterized by high dynamics, clutter, and visual monotony, construction environments pose severe navigational challenges. Current SLAM solutions impose heavy computational burdens and ...
Abstract: This study presents a 3D-printing workflow for Martian habitat components using a six-degree-of freedom (6-DoF) robotic arm, experimentally validated through scaled fabrication models for future extraterrestrial construction. The printing system was developed at METU ROMER and integrates a custom hot-melt extrusion head with a Python - Robot Operating System ...
Abstract: Designing resilient and resource-efficient habitats is essential for ensuring the safety and well-being of astronauts in the harsh Martian environment. This paper investigates optimal habitat designs capable of withstanding extreme temperatures, high radiation, and the thin Martian atmosphere, while maximizing resource efficiency and sustainability. Previous studies have explored Earth-supplied or ...
Abstract: Manual fabrication of Nail-Laminated Timber (NLT) panels relies on repetitive operations and visual judgment, which can limit consistency, productivity, and process repeatability. This study presents a comparative time-based analysis between manual and simulated robotic fabrication of full-scale NLT panels. Manual assembly was examined through video-based time studies, decomposing the workflow ...
Abstract: Thermal modeling of built environments is important for building energy monitoring, thermal performance analysis, and building envelope inspection. Robotic thermal modeling deploys a single mobile robot to collect data and construct thermal point clouds, but such systems and processes are time-consuming, expensive, and intrusive to building operations. This paper introduces ...
Abstract: Construction environments are inherently unstructured and hazardous, posing significant challenges for fully autonomous construction robotic systems. Teleoperation offers a feasible solution by enabling human operators to directly supervise and control robotic systems in the field. However, conventional teleoperation methods typically rely on control devices such as joysticks or teach pendants, ...
Abstract: The exponential growth of visual data in construction management has rendered manual image tagging inefficient and error-prone. While traditional computer vision methods offer automation, their reliance on closed-set supervised learning lacks the semantic flexibility required for dynamic, open-vocabulary site querying. To address this limitation, this study proposes a multimodal visual ...
Abstract: BIM models are typically designed separately by discipline and integrated at later stages, a process that often results in numerous clashes. Although existing commercial BIM-based clash detection tools effectively identify physical interferences, they provide limited support for semantic interpretation and decision-making, often requiring practitioners to manually analyze large volumes of ...
Abstract: Crane lifting operations strongly influence construction productivity, yet current monitoring approaches offer limited ability to interpret lifting activities beyond motion tracking. This study presents a knowledge graph approach that uses a Neo4j property graph with in-database reasoning to interpret crane lifts. A conceptual graph schema is proposed to organize fused ...
Abstract: Greenhouse gases (GHGs) contribute to the climate change intensely, in which carbon emissions play an important role. The building sector has been widely regarded as a primary contributor to carbon emissions. Steel modular construction has been considered as an innovative approach for building project delivery, particularly in temporary and emergent ...
Abstract: Temporary modular construction units are designed for repeated installation, dismantling, and redeployment across multiple sites, thereby reducing material waste and embodied emissions compared to single-use building solutions. This reuse creates a circular flow of modules, in which units return from completed projects and are redeployed to new sites. Managing these ...
Abstract: Construction industry holds a critical position in global carbon emissions and energy consumption, with its significance reflected not only in its substantial carbon footprint and energy demand share but also in its critical role in carbon reduction and sustainable development. This paper conducts comprehensive Life cycle assessment (LCA) research on ...
Abstract: With the increasing number of high-rise construction projects, the health status of tower-crane traction wire ropes is directly linked to the safety of hoisting operations. Addressing the limitations of traditional detection methods?such as manual visual inspection, magnetic induction, and ultrasonic testing, which suffer from low frequency, high subjectivity, and limited ...
Abstract: Construction projects are characterized by highly dynamic environments and concurrent multi-project execution, where safety risks continuously evolve during construction activities. Conventional construction safety management still relies on manual inspections and static reporting mechanisms, limiting real-time information integration and timely decision-making. Although recent studies have introduced artificial intelligence (AI)-based image ...
Abstract: Transportation infrastructure projects frequently experience cost changes throughout the pre-construction project development process (PDP). While transportation agencies employ contingency allowances and periodic estimate updates to manage this uncertainty, empirical understanding on how cost estimates evolve prior to contract letting remains limited. This study examines phase-to-phase cost changes across Concept, Preliminary ...
Abstract: Design for Safety (DfS) has gained increasing attention as a proactive strategy to reduce construction accidents, yet its practical implementation remains limited by manual, expertise-driven processes. Recent advances in artificial intelligence (AI), including rule-based reasoning, ontologies, natural language processing (NLP), machine learning, large language models (LLMs), and digital twins, offer ...
Abstract: The construction industry exhibits a higher fatality rate compared to other industries, making proactive safety management essential for risk reduction. Existing safety management practices have primarily focused on post-accident responses, presenting challenges in implementing preventive measures. Therefore, this study proposes a framework that integrates the Design for Safety (DfS) concept ...
Abstract: With the rapid development of robotics technology, collaborative robots have been widely used in manufacturing, construction, assembly, and logistics. While collaborative robots can improve productivity and reduce worker workload, their increased autonomy may make it difficult for workers to predict their behavior, thus causing psychological stress. However, existing research still ...
Abstract: Accident investigation is a critical component in construction safety management, yet conventional practices remain labor-intensive and time-consuming. This paper investigates the potential of Large Language Models (LLMs) to facilitate causal analysis of crane accidents by extracting and reasoning about contributing factors from narrative accident reports. Although LLMs have rapidly advanced, ...
Abstract: The management of data in complex construction projects represents a significant challenge in the design and realization of infrastructure. This paper examines the implementation of an integrated Common Data Environment (CDE) to centralize graphical and non-graphical information, aiming to optimize coordination, improve quality, and increase productivity throughout the project lifecycle. ...
Abstract: Construction robotics operates within a fragmented landscape of safety standards. While many national and international organisations have been proposing guidelines on boundaries, sensing, emergency response and risk management, these documents vary widely in how safety is articulated. However, many guidelines rely on qualitative or procedural descriptions rooted in local practices, ...
Abstract: Supply-chain characteristics are known to influence material prices, however they are often overlooked when price forecasting and analysis is done. This study seeks to address this gap by creating a framework that links machine learning model performances with inference statistic methods to understand the influence of lead-time on material price ...
Abstract: Automated progress monitoring has advanced through reality capture (images, laser scanning, SLAM/mobile mapping) and data-driven interpretation aligned to BIM/4D models. Yet many workflows still stop at percent-complete estimates and leave the schedule-update step under-specified, especially when completion depends on inspections, tests, and formal acceptance. This paper proposes a management-oriented sensing-to-scheduling ...
Abstract: Financial performance serves as the primary metric in project management to make strategic decisions based on internal and external data. While traditional manufacturing sectors utilize these indicators to examine stability, the construction industry possesses distinct characteristics defined by high material reliance and labor-intensive operations. This study employs an interpretable machine ...
Abstract: Construction Site Layout Planning (CSLP) involves the spatial arrangement of temporary facilities to optimize safety, operational efficiency, and adaptability under dynamic site conditions. Traditional optimization-based approaches, while effective at improving objective performance, often converge toward a narrow set of solutions and overlook diverse yet feasible alternatives that are critical for ...
Abstract: Construction, renovation and demolition activities generate 30-40% of global solid waste. Yet, buildings are rarely designed for repair and reuse. Traditional Japanese timber joinery offers a historically grounded counter-model. Daimochi tsugi is a shear-resistant load-bearing scarf joint used in the roof structures of temples, shrines and wagoya buildings since the ...
Abstract: Construction site safety management is challenged by the difficulty of accessing and applying safety procedures during on-site activities, particularly in dynamic and multicultural environments. While Large Language Models offer new opportunities for information retrieval, their use in safety-critical construction contexts raises concerns related to the risk of hallucination, source reliability ...
Abstract: Building Information Modeling (BIM) is an innovative digital approach in the construction industry that enhances conventional practices such as cost estimation and scheduling. It employs a concise 3D model containing geometry to depict actual conditions and produce precise quantity take-offs and time-phased visualizations. However, there are still limited practices for ...
Abstract: While accurate planning is fundamental to effective construction project management, historical data and existing planning tools remain insufficient to inform manufacturers' decision-making in off-site construction (OSC) production facilities. Although surveys are commonly used to collect industry knowledge, most prior efforts have focused on on-site construction processes, and their findings have ...
Abstract: Carbon markets offer a promising path to unlock carbon-emission reduction (CER) potential in the construction industry; however, the industry often lacks sufficient incentives to adopt CER activities. This study proposes a blockchain-based consensus mechanism, namely Proof-of-Carbon-Sold (PoCS), to incentivize CER of the construction industry in the carbon markets by linking ...
Abstract: Construction and demolition waste (C&DW) management is central to urban renewal and sustainable construction, yet current intelligent solutions often suffer from fragmented multimodal data, weak knowledge linkage, and limited policy interpretability. Image recognition models can classify C&DW materials but cannot integrate regulatory clauses, disposal processes for explainable reasoning, while text ...
Abstract: Urban green infrastructure is increasingly recognised as a critical component of sustainable and resilient cities. However, planning and construction decisions often fail to capture the coupled impacts of environmental performance and human behaviour in a systematic and quantitative manner. This paper presents an AI-enabled Climate-Health Nexus (CHN) framework that integrates ...
Abstract: Human-Robot Collaboration (HRC) is increasingly explored in construction. However, the health and safety implications of HRC remain only partly understood, particularly ergonomic risks leading to musculoskeletal disorders (MSDs) of construction workers in HRC. This review synthesized existing research on health and safety in construction HRC and examined ergonomic risks leading ...
Abstract: Ergonomic risk assessment helps improve safety and reduce injuries in construction work. The Rapid Entire Body Assessment (REBA) method is widely used, but it is usually done manually. This makes it slow, subjective, and hard to apply, especially on large construction sites. Though vision-based methods do exist, many tend to ...
Abstract: Team Situation Awareness (TSA), a collective understanding of changes in the surroundings, is critical in human-human and human-robot teaming in the futuristic construction site. Current measuring techniques fall short in capturing continuous and real-time data in realistic settings. Psychophysiological signals, including visual attention (eye-tracking) and neural response (prefrontal cortex ...
Abstract: Worker stress is a primary contributor to accidents on construction sites. While wearable biosensors enable continuous stress monitoring, existing approaches treat stress either as a binary problem or a flat level-classification task, overlooking the distinction between two stress appraisals (i.e., challenge and threat) and the hierarchical nature of the stress ...
Abstract: Human-robot collaboration (HRC) is increasingly adopted in prefabricated construction to improve productivity while maintaining flexibility. However, most existing HRC planning approaches remain efficiency-oriented or rely on simplified fatigue assumptions that lack strong empirical grounding in prefabrication work, which is characterized by frequent switching between physical execution and cognitive monitoring. This ...
Abstract: Roofers often work on sloped surfaces and in awkward postures, which increases their risk of falls and injuries. In recent years, drones have been increasingly used in construction, but their presence may distract workers. Such distractions can compromise balance, which is especially critical for roofers working in challenging conditions, where ...
Abstract: The construction industry faces challenges of low productivity, high physical workload, and workforce shortages, especially in repetitive tasks like wall and ceiling spraying. This study evaluates the impact of robotic assistance on ergonomics and process efficiency using a digital process twin combining a digital human model (DHM) with motion-captured movements. ...
Abstract: Accurate ergonomic risk assessment is critical for occupational safety in construction. However, the widely used observation-based tool Rapid Entire Body Assessment (REBA) may suffer from low sensitivity arising from stepwise score assignment and limited accuracy due to joint angle categories defined without the support from muscle activity data. To ...
Abstract: Natural language interfaces (NLIs) have been explored as a way to lower the entry barrier to interacting with infrastructure digital twins (DTs) by reducing the need for extensive prior training. However, most existing NLIs rely on manually configured parsing logic based on fixed and predefined data schemas, making them difficult ...
Abstract: Fire hazards on construction sites often involve complex visual and contextual cues that are difficult to detect and interpret. While immersive virtual reality (IVR) and eye-tracking technologies have been increasingly adopted for construction hazard recognition, limited work has examined how fire hazard recognition failures arise. Situational awareness (SA) theory provides ...
Abstract: Human-robot collaboration (HRC) has emerged as a promising approach to improve productivity and safety in construction. However, most existing construction task scheduling methods are predominantly deterministic and fail to account for the significant uncertainty in human task execution caused by factors such as skill variability and fatigue. Such limitations can ...
Abstract: Work-related Musculoskeletal Disorders (WMSDs) are a leading cause of worker attrition in construction, yet countermeasures remain largely reactive, relying on post-hoc observations. Traditional WMSD simulation tools based on pre-defined motion templates are suitable for static environments but remain limited in handling selected sources of condition variability in representative construction-style tasks. ...
Abstract: This research proposes a framework that enables construction workers to control robots using natural language, without needing to write code. The construction industry faces two problems: not enough workers, and robots that are too hard to use. Surveys show that technical complexity and lack of training are major barriers to ...
Abstract: In the construction industry, research on Human-Robot Collaboration (HRC) is advancing to leverage the complementary strengths of both humans and robots. While research on HRC in architecture is progressing, most studies focus on task partitioning. There remains a lack of research on co-manipulation, where humans and robots handle the same ...
Abstract: Designing adaptive human-robot collaboration (HRC) systems for timber construction is challenging due to the difficulty of encoding tacit construction knowledge and material variability into computational models. This paper presents a material-aware probabilistic framework for light-frame timber assembly. A lightweight timber construction ontology is extracted from construction manuals and empirical observations ...
Abstract: Persistent labor shortages and safety exposure motivate close-contact human-robot collaboration (HRC) for construction assembly. Training proactive collaborative robots for assembly, however, commonly depends on largescale demonstrations and extensive sensing, which increases data-collection cost and limits portability across variable workspaces. This paper develops a standard operating procedure (SOP)-grounded vision-language-action (VLA) policy. ...
Abstract: AI-assisted voice control is increasingly used in hands-busy human-robot collaboration, yet its effects on trust and task-grounded behaviours remain rarely explored. This study evaluated AI-assisted voice control in a scaffold-assembly handover task, comparing it against a No-voice Baseline while keeping robot paths and task layout constant. Workload (NASA-TLX), safety, performance, ...
Abstract: The Architecture, Engineering, and Construction (AEC) industry continues to lag behind other sectors in robotics adoption. A persistent barrier is conceptual ambiguity around what constitutes Human-Robot Collaboration (HRC) in AEC, where definitions are often inherited from manufacturing and do not translate well to dynamic, unstructured work sites. To address this ...
Abstract: This paper establishes a performance evaluation framework for Remote-Controlled Tower Cranes (RCTCs), which potentially addresses the growing need for standardized, safe and efficient lifting operations in high-rise building works and large-scale civil infrastructure sites. Specifically, four performance metrics are defined: (1) lifting productivity, (2) system responsiveness, (3) operator wellness and ...
Abstract: This study aims to explore how crowdsourcing mechanisms can address the limitations of traditional disaster monitoring systems regarding coverage and timeliness. Targeting the general public, this research leverages the features of instant messaging software and online forms to design disaster data collection interfaces. Furthermore, various platforms?including LINE, Google Forms, and ...
Abstract: The continued digitalization of construction practice, including the expanded use of BIM, automation, and data-centric workflows, has increased the demands placed on construction management education. As curricula adapt to these developments, both instructors and students face growing challenges in organizing and interpreting course content during independent study. Traditional slide-based review ...
Abstract: Wider adoption of robots in construction is becoming increasingly feasible as sensing, control, and artificial intelligence (AI) technologies advance. However, the inherent complexity and variability of construction activities mean that full automation remains challenging, making human-robot collaboration (HRC) a more practical approach for structural assembly. Effective HRC requires workers to ...
Abstract: The growing use of unmanned aerial vehicles (UAVs) in built environments has created a need for simulation tools that support realistic and systematic human-UAV interaction (HUI) research. However, many existing HUI studies still rely on simulation setups that simplify flight control, onboard computation, sensing, or environmental conditions, limiting the ecological ...
Abstract: Access to active construction sites is often limited in construction engineering education due to safety, logistical, and scheduling constraints. To address this issue, a 1:50 scale physical construction site model was designed as a pedagogical tool based on a real site layout. The model represented a building under construction and ...
Abstract: Work-related musculoskeletal disorders (WMSDs) remain a major safety issue in construction, particularly for scaffolding workers who are repeatedly exposed to awkward and constrained postures. Conventional ergonomic assessment methods such as REBA and RULA rely on manual observation and static posture sampling, which limits their applicability for continuous monitoring in dynamic ...
Abstract: This study investigates the safety compliance of pedestrian within smart traffic systems using a multimodal data approach. By integrating behavioral records, questionnaires, eye-tracking metrics, and electroencephalography signals, the study examines the cognitive mechanisms underlying pedestrian decision-making during street crossing. The research utilizes structural equation modeling to assess how environmental factors, ...
Abstract: Manual construction scheduling is prone to errors and suboptimality. To this end, there is growing attention to reinforcement learning (RL)-based approaches for automated construction scheduling due to their superior time efficiency and solution quality compared to metaheuristic algorithms. However, developing suitable reward functions for training RL agents to solve real-world ...
Abstract: The construction industry remains one of the high-risk industries worldwide. Improvement of construction worker health and safety requires a safer interaction between workers and construction tasks including materials used, tools adopted, and protective systems applied. Construction safety research has focused on human factors such as worker behavior, workload, and safety ...
Abstract: Language discordance remains a persistent barrier to effective safety communication on U.S. construction sites, particularly for time-critical hazards such as heat stress. Existing digital safety tools and translation applications rely on scripted content, text-based interfaces, or cloud connectivity, limiting their effectiveness in dynamic, multilingual job-site environments. This paper presents a ...
Abstract: Steel construction is a widely adopted building method that relies on durable, high-strength materials to achieve efficient and reliable structures. Artificial Intelligence (AI) technologies have the potential to enhance efficiency, precision, safety, and overall productivity within steel fabrication, a process that remains largely dependent on manual operations. Despite promising advancements, ...
Abstract: To support net-zero goals, greenhouse gas emissions from the civil engineering and construction industries are receiving increasing attention. Assessing and reducing embodied carbon during the design phase is crucial for achieving these goals. However, many commercial BIM-based LCA tools rely on international databases, which may not accurately reflect the material ...
Abstract: Prefabricated Prefinished Volumetric Construction (PPVC) is increasingly adopted under Design for Manufacturing and Assembly (DfMA), yet the geometric conformity of structural components remains a key challenge?particularly when alternative fabrication methods such as 3D concrete printing (3DCP) are considered. Due to layer-wise deposition and printhead kinematics, 3DCP wall specimens may exhibit ...
Abstract: This paper proposes and evaluates an automated BIM-Blockchain workflow to certify BIM design changes as verifiable digital evidence, aligned with ISO 19650-oriented traceability indicators. First, baseline traceability was assessed through a two-layer checklist (technical and operational) applied to a real residential BIM project developed in Autodesk Revit. Then, an end-to-end ...
Abstract: Construction schedules serve as one of the primary doc uments for encoding project intent in the planning phase, yet their unstructured, inconsistent, and ambiguous term renders them inaccessible to processing systems. These in stances present a critical barrier to automation for construc tion actions, as processing requireexplicit, machine-readable representations of ...
Abstract: Low-cost quadruped robots offer an affordable platform for research and education, yet achieving reliable locomotion remains challenging due to servo-driven actuation, sensing noise, and mechanical uncertainty. Although reinforcement learning has shown strong performance on high-end quadrupeds, its effectiveness on low-cost platforms is not well understood. This paper presents a ...
Abstract: Registration between as-designed BIM and as-built point clouds is a critical technology in digital construction. However, challenges such as partial overlap and substantial noise limit the effectiveness of traditional methods, particularly in large-scale building scenarios. To address these limitations, this study proposes a building structure matching method based on a ...
Abstract: Existing camera-based crane hand gesture recognition methods remain vulnerable to background variation, illumination changes, and partial hand visibility, which limits their robustness in real construction environments. The study presents a multimodal crane hand gesture recognition framework that integrates helmet-mounted binocular vision with wearable sensing. Dual-view visual inputs are processed using ...
Abstract: The high frequency of human activity in residential buildings has led to a continuous occurrence of fire accidents, which has become a prominent threat to public safety. Addressing the bottlenecks of reliance on manual experience, insufficient coverage, and low efficiency in fire scenario settings for performance-based fire design (PBFD) of residential ...
Abstract: Human-robot collaboration (HRC) in construction requires robots to understand worker intent and predict subsequent actions for proactive assistance. However, existing methods are primarily limited by closed-vocabulary classification, hindering their adaptability to diverse construction tasks. These approaches are also easily compromised by site environmental factors like poor lighting and severe occlusions. ...
Abstract: Urbanization-driven construction and demolition activities have caused a global surge in construction and demolition waste, posing severe environmental threats. As a high-density metropolis with intensive construction, Hong Kong faces acute construction and demolition waste management challenges, such as low recycling rates and illegal dumping. Accurate estimation of construction and demolition ...
Abstract: This study proposes a quantitative framework for assessing the cognitive development of tower crane operators by integrating high-fidelity Virtual Reality (VR) simulation with the SEEV (Salience, Effort, Expectancy, Value) attention model. Forty participants completed a longitudinal training protocol incorporating a strict restart-on-error mechanism to simulate safety-critical operations. Eye-tracking data were ...
Abstract: Reliable fault diagnosis of high-speed train bearings remains challenging due to strong noise interference and pronounced domain discrepancies between laboratory test conditions and real in-service operating environments. To address the degradation of diagnostic performance under cross domain scenarios, this study proposes a physics informed adversarial transfer learning framework for intelligent ...
Abstract: Conventional synchronization between physical and structural BIM representations typically relies on merging adjacent nodes by altering element locations, which causes challenges for bidirectional interoperability. This discrepancy arises from the fundamental domain difference between high-fidelity physical representations and frame-based analytical models, often necessitating labor-intensive manual corrections. To address this challenge, this ...
Abstract: Artificial Intelligence (AI) is increasingly being adopted across various areas of construction projects. Previous studies have identified applications of AI in construction, including improving construction site safety through real-time monitoring and predictive analytics; optimized planning and scheduling; automating construction activities using robotics; identifying construction risks and potential delays; improving allocation ...
Abstract: Effective knowledge sharing and reuse across construction projects are frequently impeded by various project conditions and user requirements. Semantic web technologies (SWTs) provide a structured method for managing reusable knowledge in such dynamic environments. Nevertheless, the broader adoption of SWTs has been constrained by the limited availability of skilled knowledge ...
Abstract: Construction projects are often challenged by time and cost overruns due to the poor quality of their manually constructed schedules. This is why recent research has been exploring how to leverage the Information Model of the to be constructed Building (BIM) to automatically produce a high-quality schedule. The major impediment ...
Abstract: Urban Taiwanese townhouses are typically narrow, deep, and tightly spaced, with limited façade openings. These conditions restrict wind access and cross-ventilation, amplify solar-driven heat accumulation, and increase reliance on air conditioning. Although termite-mound-inspired ventilation has been implemented in commercial buildings, its effectiveness for high-density residential townhouses remains underexamined. This study ...
Abstract: Accurate as-built BIM models are essential for construction monitoring, Quality Control (QC), and Facilities Management (FM). However, current Scan-to-BIM workflows struggle in reconstructing accurate as-built models. This paper presents the inverse kinematics -based iterative closest point (ICPIK) method for automated parametric modelling of deformed elements without manual intervention. Experimental ...
Abstract: Automated understanding of complex construction hazard scenarios is essential for intelligent site safety management. However, existing computer vision and image captioning approaches often lack interpretability and semantic consistency required for engineering decision support. This study proposes a Hazard Ontology-Guided Attribute-Based Model (HOGAM) for automated construction hazard description. The framework decouples ...
Abstract: This paper presents a facade-level Scan-to-BIM pipeline from terrestrial laser scanning (TLS) data for energy renovation applications. The workflow combines image-based edge extraction, geometric processing, and procedural BIM modeling. A Point Cloud Edge (PCE) module extracts facade contours from panoramic TLS projections, which are then processed through plane detection, clustering, ...
Abstract: Precise segmentation of shield tunnel linings is a prerequisite for structural health monitoring and digital twin construction. However, existing methods are heavily relying on extensive annotated datasets, which are scarce in tunneling scenarios. To address this, we propose a novel zero-shot framework for the adaptive recognition and instance segmentation of ...
Abstract: Reinforcing bars (rebars) are essential for resisting tensile forces and ensuring structural safety in reinforced concrete systems. However, current rebar inspection practices rely on manual and sampling-based measurements, leading to human errors and limiting inspection coverage. Although automated 3D and LiDAR-based inspection systems have been explored, their adoption rate remains ...
Abstract: Construction robotics has a significant potential to automate a wide range of construction tasks, but it depends on a structured pipeline that links high-level design representations to concrete robotic actions. Robotic bricklaying is particularly well-suited to automation because it consists of repetitive, predefined steps. However, it also introduces process-planning ...
Abstract: Scaffold inspection is critical for ensuring construction safety. However, scaffolds are typically dense, multilayered, and continuous structures, making them difficult to analyze automatically using traditional geometric methods. In this paper, a framework for automated scaffold component recognition and joint-level understanding is proposed by leveraging Vision-Language Models (VLMs) and 3D Gaussian ...
Abstract: Fire safety compliance checking is essential for BIM-based design review, yet it remains largely manual in practice, leading to substantial time costs. Although automated checking approaches (e.g., NLP-based methods) have been explored, two major challenges remain: (1) in practice, operationalization is largely limited to semantic requirements, whereas clauses involving geometric ...
Abstract: To address the limitations of the traditional Fast Point Feature Histogram (FPFH) in heterogeneous building point cloud registration?including over-reliance on geometric information, neglect of color information, and susceptibility to lighting variations?this paper proposes a color-enhanced local feature fusion framework. First, the RGB color space is transformed into the HSV space, ...
Abstract: With advancements in large language models (LLMs), there has been growing attention to automatic construction specification reviews, enabling faster and more standardized assessments. However, general-purpose LLMs' effectiveness has been bounded when applied to construction contexts, because they lack necessary domain-specific vocabulary and knowledge. Therefore, we propose a domain-specialization LLM pretraining ...
Abstract: The construction of operating rooms involves multiple coordinated procedures. In practice, time and cost estimation are often uncertain. Additionally, project control still relies heavily on the manual interpretation of 2D drawings, resulting in inconsistencies between the design intent and on-site execution. There is one other thing that is important for ...
Abstract: Full-scale 3D concrete printing for horizontal structural components is challenging due to large unsupported spans, early-age material instability, and geometric limitations of extrusion-based processes. A practical solution is to fabricate modular slab units by printing only the structural ribs of the slab while casting the top and bottom layers conventionally ...
Abstract: Quality control is critical for off-site construction, but it still depends heavily on individual expertise and manual inspection. The management and coordination of heterogeneous and fragmented knowledge among the stakeholders involved in quality control of off-site construction remains challenging. To address this knowledge gap, this paper proposes QuosRA, a novel ...
Abstract: Highway project information delivery using building information modelling (BIM) often suffers from inefficiencies when converting 3D Industry Foundation Classes (IFC) models into network representations that conform to agency-specific asset data standards. These inefficiencies primarily arise from inconsistent object classification and mismatches in modelling granularity. To address this challenge, this paper ...
Abstract: Flood damage estimation for residential buildings is a critical component of post-disaster recovery planning, budgeting, and construction management. Conventional depth-damage curves (DDCs), which remain the standard tool in flood risk assessment, rely on deterministic and monotonic assumptions that fail to capture the substantial variability observed in real-world flood losses. As ...
Abstract: Building permit applications usually consist of multiple heterogeneous, unstructured, and non-machine-readable data. To efficiently check and process applications, relevant information must be extracted, and documents must be classified and linked to the application and process steps in the review. To address this, this study presents an end-to-end method for generating ...
Abstract: With the growing use of artificial intelligence in architectural design, BIM models now need to record not only geometric shapes but also process information and external knowledge for intelligent design and optimization. This paper introduces an enhanced semantic IFC (Industry Foundation Classes) model to transmit process and semantic data during ...
Abstract: Geometric measurement is critical for Modular Construction (MC) after fabrication because it helps prevent dimensional mismatches and safety risks that might otherwise be discovered only during on-site assembly. Terrestrial laser scanning (TLS) provides dense 3D measurements, but automated measurement in cluttered MC scenes remains challenging without robust semantic understanding and ...
Abstract: Crack segmentation is a critical yet challenging task in structural health monitoring, as cracks typically appear as thin, low-contrast structures that are easily confounded by complex background textures. Although recent deep learning methods have substantially improved crack segmentation accuracy in conventional RGB imagery, they often overfit dataset-specific textures, resulting in ...
Abstract: Falls from height (FFH) remain a leading cause of construction injuries and fatalities, necessitating proactive Design-for-Safety (DfS) approaches. While Building Information Modeling (BIM) facilitates safety checks, existing Automated Rule Checking (ARC) models rely on rigid, manual rule-coding and lack the spatial reasoning required to interpret complex safety regulations. This paper ...
Abstract: Assembly Sequence Planning (ASP) is a critical decision-making problem in construction automation, directly influencing productivity, cost efficiency, constructability, and system robustness in prefabricated and modular construction systems. The ASP problem is inherently NP-hard (Non-deterministic Polynomial-time hard) due to the factorial growth of feasible assembly sequences as product complexity increases. While ...
Abstract: Generating useful BIM geometry from laser scan data remains a major bottleneck, especially for hidden Mechanical, Electrical, and Plumbing (MEP) systems in cluttered timber construction. Pipes and wires are thin, fragmented, and often partially occluded, making geometric reconstruction highly sensitive to manually tuned parameters. This paper presents Cognitive Fitting, an ...
Abstract: Tunnel engineering specifications are critical normative documents for compliance verification and decision support. Existing text-based retrieval methods do not sufficiently exploit the hierarchical structure between clauses, making it difficult to reliably locate evidence and prioritize core clauses. This paper proposes a clause-level retrieval and evidence localization method for tunnel specifications ...
Abstract: Ageing bridge infrastructure requires scalable, accurate, and minimally intrusive inspection methods that can reveal subsurface conditions while producing outputs compatible with digital engineering workflows. Muon Flux Technology (MFT) enables passive, non-destructive internal imaging of reinforced concrete elements, producing high-value geometric outputs that can support infrastructure inspection and asset management. A ...
Abstract: Stray animals colliding with moving vehicles is one of the most serious road safety hazards that exists today. Many of these collisions occur at roads that run directly next to important urban spaces like institutional corridors, where stray animals often invade the highways. Since the use of physical barriers and ...
Abstract: Construction robots require component-level task specifications that include per-unit dimensions, 6D poses, and assembly sequences respecting construction constraints. However, existing BIM-based robot task planning pipelines rely on either ad-hoc heuristic decomposition or exhaustive pre-modeling, both of which are brittle under design changes and fail to capture the assembly semantics encoded ...
Abstract: Engineering drawings remain a critical medium for conveying design intent in architectural and construction projects. Despite the increasing adoption of Building Information Modeling (BIM), two-dimensional drawings are still widely used in practice, particularly during early design stages and interdisciplinary coordination. The interpretation of such drawings largely relies on manual processes, ...
Abstract: Recent advances in three-dimensional (3D) generative Artificial Intelligence (AI) hold significant potential for automating architectural design. However, existing methods cannot effectively process Industry Foundation Classes (IFC) data due to its complex structure. There is a pressing need for a unified representation that preserves the object-level spatial information for generative tasks. ...
Abstract: Flooding can severely disrupt highway networks, causing substantial mobility loss and delayed recovery. Existing studies that quantify flood impacts on transportation networks primarily focus on structural or functional features, largely overlooking the role of socioeconomic vulnerability. Additionally, how socioeconomic vulnerability influences flood impacts and recovery prioritization across traffic-demand groups remains ...
Abstract: Alignment between Building Information Models (BIM) and as-built representations is essential for automated construction progress monitoring. Existing pipelines that rely on images often require manual selection of correspondences or additional site instrumentation. Recent advances in differentiable scene representations, especially 3D Gaussian Splatting (3DGS), open new possibilities for automatic global BIM ...
Abstract: Underground working environments are inherently hazardous, with explosion risks posing a persistent threat. Traditional static risk assessments struggle with dynamic conditions. This paper presents a novel real-time explosion risk prediction framework integrating Internet of Things (IoT) sensor data, Long Short-Term Memory (LSTM) networks, and Bayesian Networks (BNs). The system monitors ...
Abstract: Recent construction sites increasingly adopt real-time safety monitoring that applies rule-based logic to frame-level object detections. Yet such approaches provide limited evidence grounded in relationships among workers, PPE status, actions, locations, and hazards. Construction videos are highly variable, so detections fluctuate over time (temporal inconsistency); even in the same situation, ...
Abstract: Representation learning for 3D scenes in the architecture, engineering and construction domains is becoming important, as it facilitates Scan-to-BIM, 3D perception and robotic deployment in built environments. This study presents NUS3D, a dataset that consolidates BIM-synthetic, mobile-scan and terrestrial laser scan (TLS) point clouds for building interiors. Originated in 2022, ...
Abstract: Modular construction offers significant efficiency benefits over traditional methods, yet its production systems remain vulnerable to workflow variability caused by product customization and manual operations. While buffering is a standard strategy to mitigate such variability, the unique constraints of modular factories (e.g., specifically the large physical footprint of modules and ...
Abstract: Reinforcement detailing quality in reinforced concrete structures is directly associated with structural safety and project costs. However, manually cross-checking structural construction drawings against rebar schedules remains inefficient and highly susceptible to human error. Although demand for digital transformation in construction has become increasingly urgent, achieving high-precision automated information extraction continues ...
Abstract: Occurrences of blockages in stormwater drainages could be a significant issue, especially in urban areas, due to various challenges for infrastructure operation and maintenance. Predicting blockage occurrences is a complex problem that involves numerous parameters and an adequate understanding of their impacts, interactions and the nonlinear behaviours of certain parameters. ...
Abstract: Electrical system modeling in BIM environments still relies heavily on manual interpretation of 2D CAD drawings, which can propagate circuit and panelboard assignment errors and increase rework. This paper presents an AI-driven, a parametric CAD-BIM workflow that automates outlet-to-circuit assignment and circuit-to-panelboard connections while maintaining semantic traceability. CAD plan views ...
Abstract: Automated point cloud understanding is critical for modern digital construction workflows, yet existing datasets focus predominantly on architectural elements, leaving complex mechanical, electrical, and plumbing (MEP) systems severely underrepresented. This paper introduces Industrial3D, the first large-scale, high-fidelity dataset specifically designed for industrial MEP scene understanding. Industrial3D comprises over 610 million ...
Abstract: Construction robotics has significant potential to improve the efficiency, quality, and safety of structural assembly, yet its practical deployment is constrained by the limited availability of online task sequence optimization methods. This study focuses on gantry robots performing connection tasks in floor frame assembly under limited nail magazine capacity. To ...
Abstract: Semantic understanding of indoor building environments is essential for digitalization, facility management, robotics, and human-computer interaction. However, existing workflows typically decouple data acquisition, semantic interpretation, and 3D reconstruction into offline processes, limiting on-site verification and interaction. This paper presents an extended reality assisted framework for semantic understanding and digitalization of ...
Abstract: Reliable and timely construction progress monitoring remains challenging due to the dynamic, spatially complex, and visually cluttered nature of jobsite environments, where conventional inspection- and report-based approaches often yield subjective and intermittent assessments. This paper maps the adoption of artificial intelligence and computer vision for construction progress monitoring through a ...
Abstract: Tall renewable energy infrastructures are complex structural systems comprising numerous components. Under earthquakes, their designated energy-dissipating components may exhibit intricate material and geometric nonlinearities that require high-fidelity finite element (FE) modelling, whereas system-level seismic assessment typically demands extensive nonlinear time-history analyses, making uniformly refined modelling computationally prohibitive. This paper presents ...
Abstract: The digital transformation of existing building stock through Building Information Modeling (BIM) remains a significant goal and challenge in the Architecture, Engineering, and Construction (AEC) industry. This paper presents an automated Scan-to-BIM framework specifically designed for IFC-compliant roof model generation from Unmanned Aerial Vehicle (UAV) photogrammetry data. The proposed method ...
Abstract: Ballastless track slab systems are increasingly adopted in metro and high-speed rail projects due to their superior geometric stability, reduced maintenance requirements, and long-term performance. Construction of such infrastructure is more mechanised, driven by stringent requirements on construction speed, dimensional precision, and repeatability in processes such as material handling, batching, ...
Abstract: Accurate indoor self-localization remains a fundamental challenge in environments where GPS signals are unavailable, such as construction sites, underground facilities, indoor spaces, and tunnels. This paper presents a smartphone-only indoor self-localization and navigation system that uses only sensors embedded in standard smartphones. The proposed system utilizes smartphone cameras and inertial ...
Abstract: In construction and operational stages of buildings and in computer vision and construction robotics, alignment of LiDAR point clouds to a building's 3D coordinate axes is a must-have preprocessing step. However, existing point cloud processing tools handle a point cloud in a single batch, which can lead to high memory ...
Abstract: Reliable and rapid energy forecasting is essential for early-stage decision-making in office building design, yet data-driven modeling in Taiwan is hindered by the lack of region-specific datasets that reflect local building typologies and climates. This study addresses this gap by introducing TOBE (Taiwanese Office Building Energy), a BIM-derived dataset developed ...
Abstract: Generating building digital twins from UAV imagery via 3D reconstruction is significant. However, reflective glass façades remain particularly challenging: their high-frequency, view-dependent appearance often causes photogrammetry to yield distorted geometry and corrupted textures. While Gaussian Splatting supports view-dependent rendering, its spherical-harmonics color model mainly captures low-frequency components and fails to ...
Abstract: Railway sleepers form the structural interface between the rails and the ballast, governing track geometry stability through load transfer and positional restraint, making them a critical track component. Regular monitoring of sleeper condition and spacing contributes to maintaining track condition and supports safe railway operation for goods and passenger traffic. ...
Abstract: Basement construction projects often experience low productivity, schedule delays, cost escalation, safety risks, and high embodied carbon. These challenges largely arise from uncertainty in ground conditions and construction-stage soil-structure interaction, which cannot be fully resolved during design. The observational method offers a structured means of managing this uncertainty by updating ...
Abstract: This paper presents a data-driven approach to identifying candidate spatial signals of interface dependencies in large-scale civil infrastructure projects by leveraging empirically observed co-change as an observable proxy for coordination-related constraints. Using tender and as-built 3D models from an expressway project, we segment the models into analysis objects using category-specific ...
Abstract: The shift toward Off-Site Construction (OSC) is constrained by the complexity of High-Mix, Low-Volume (HMLV) production, where geometric variability and reliance on manual labour create highly stochastic processing times. In these Dual-Resource Constrained (DRC) systems, workers act as flexible assets who move across stations to relieve bottlenecks?behaviour that static scheduling ...
Abstract: Crane operations pose significant safety risks on construction sites, particularly when multiple mobile cranes operate concurrently in confined and dynamic environments. This paper presents a vision-based framework for real-time multi-crane safety monitoring using Scene Coordinate Regression (SCR). By directly regressing 3D scene coordinates from monocular images, the proposed approach enables ...
Abstract: Modular construction projects rely heavily on large trucks to transport prefabricated modules, making the design of on-site haul roads critical for ensuring safe and efficient logistics operations. However, the manual design process suffers from inefficiencies. To address this issue, this study proposes MiCRoadDiff, a diffusion model-based intelligent approach for modular ...
Abstract: Off-site construction offers significant potential to improve productivity and quality in the Architecture, Engineering, and Construction (AEC) industry; however, the practical adoption of Design for Manufacturing and Assembly (DfMA) remains limited due to the lack of explicit integration between product-focused Building Information Modeling (BIM) data and construction process knowledge. While ...
Abstract: Deep learning-based computer vision in construction has often assumed that more training data leads to better performance. This quantity-driven approach overlooks the detrimental effects of redundant, noisy, or mislabeled samples that can degrade model accuracy while inflating computational costs. To overcome this limitation, we propose and test an alternative hypothesis: ...
Abstract: Robotic total stations have transformed surveying and construction mapping through precise, efficient, and automated measurements. These instruments integrate a theodolite, which measures horizontal and vertical angles, with an electronic distance measurement (EDM) unit to determine distances, allowing accurate 3D measurement of observable points. Traditionally, users manually aimed the total station ...
Abstract: The increasing volume and heterogeneity of data generated during building inspections pose significant challenges to maintenance management (MM). Although digital technologies are frequently discussed in the literature, their integrated applications to support decision-making in maintenance are still scarce. This study proposes an integrated data visualization workflow to support building MM, ...
Abstract: Automatic concrete spalling detection on bridge surfaces using unmanned aerial vehicles (UAV) is essential for improving inspection efficiency and operational safety. However, instance segmentation models employed in UAV-based inspections are sensitive to environment-related variations in image appearance. In practice, inspection data are usually collected under limited conditions, which reduces model ...
Abstract: The construction industry's transition to off-site prefabrication requires rigorous workflow monitoring to fully realize the efficiency gains of lean manufacturing. However, measuring the variations in cycle times for components at the workstation level, such as prefabricated wooden panels, remains a challenge; traditional manual tracking is labor-intensive and error-prone, whereas supervised ...
Abstract: This paper presents a comparative study of three progressive zero-shot approaches for automated recognition of construction activities to estimate productivity using vision-language models (VLMs). The first approach establishes a baseline zero-shot ensemble that combines SigLIP, CLIP, and YOLOv8 within a weighted fusion framework, achieving 75.89% accuracy across eight concurrent construction ...
Abstract: Classification of 3D point clouds is fundamental to semantic labelling and scene understanding in bridge inspection and management. However, accurate classification of bridge components remains challenging due to their complex geometries and high inter-class geometric similarity. To address these challenges, this study explores the potential of incorporating structural knowledge into ...
Abstract: This paper presents FloorPlan2Nav, a framework that transforms architectural floor plan images into topological navigation graphs serving as prior maps for mobile robots. The framework combines deep learning-based semantic segmentation with contour-based instance extraction, where majority voting within geometrically accurate contours corrects boundary errors in pixel-wise predictions. A connectivity graph ...
Abstract: Road construction consumes large quantities of virgin aggregates and asphalt, while circular-economy (CE) strategies, such as using reclaimed asphalt pavement (RAP) and recycled aggregates, remain underutilized due to quality uncertainties, logistical constraints, and limited digital integration. Existing simulation tools are predominantly linear and single-site, which prevents the realistic evaluation of ...
Abstract: BIM design workflows involve complex, multi-step interactions with professional software, which limit the direct applicability of large language models (LLMs). This paper presents a BIM design assistant driven by LLM agents, enabling natural-language-based assistance for BIM design and visualization tasks through reliable tool invocation. First, a two-layer Revit interface function ...
Abstract: Onshore wind power plays a critical role in sustainable energy development. However, its deployment is often constrained by complex construction-stage logistics, particularly the planning and design of access roads. During the early-stage evaluation of onshore wind projects, construction road layout design and earthwork estimation remain labor-intensive due to terrain and ...
Abstract: The digitization of building codes is critical for Automated Compliance Checking (ACC)within the BIM paradigm, yet faces challenges such as unstructured text,semantic ambiguity,and deep logical nesting. Traditional hard-coding lacks flexibility,while reliance solely on Large Language Models (LLMs)is prone to hallucinations and offers limited BIM integration.This study proposes a multi-stage framework ...
Abstract: Adoption of Artificial Intelligence (AI) is growing across construction industry, but not evenly. While many companies are exploring AI for workflows like document management, AI adoption for scheduling particularly remains limited. On the other hand, industry demands adaptive scheduling due to its dynamic nature. Although traditional scheduling approaches have made ...
Abstract: Design for manufacturing and assembly (DfMA), extended through more generic Design for X (DfX) is an important component for realising the benefits of modular and industrialised construction. This requires balancing heterogeneous objectives derived from diverse, often fragmented data sets. In the absence of a unified data model, designers often resort ...
Abstract: In the field of Facility Management (FM), the increasing complexity of operations and the demand for ESG compliance require strategic, lifecycle-oriented data management. However, current practices often suffer from information fragmentation and a lack of integration between spatial and FM data. This study addresses these issues by developing a cloud-based ...
Abstract: Construction code retrieval is a recurring bottleneck in AEC compliance workflows, where decisions require clause-addressable evidence and cross-discipline reference completion under incomplete project attributes. A clause-centric retrieval-augmented generation (RAG) pipeline treats retrieval as auditable evidence discovery and enforces an evidence-bounded selection contract: outputs may cite only clause identifiers from the ...
Abstract: Rapid informal urban expansion in high-slope terrains presents a critical challenge for disaster risk management, particularly where foundational digital documentation is absent. This study proposes an automated methodology for quantifying urban growth and geodynamic hazard exposure by integrating UAV photogrammetry with a Scan-to-BIM workflow. By processing high-resolution point clouds through ...
Abstract: Structural Health Monitoring (SHM) systems installed on bridges are increasingly being used to analyze traffic parameters. This includes estimating the gross vehicle weight and may be referred to as Weigh-in-Motion (WIM) methods. However, current research relies predominantly on numerical analyzes, whereas available field experiments are typically performed under quasi-controlled conditions. ...
Abstract: Heat-related illnesses (HRIs) pose a persistent threat to construction workers' safety. Effective risk assessment enables early detection of hazardous conditions and timely prevention of illness, yet incident logs and narrative reports are seldom leveraged for computational risk analysis. This research develops a data-driven framework to (i) identify HRIs from routine ...
Abstract: Work-related musculoskeletal disorders (WMSDs) remain a critical cost driver in construction safety. While computer vision enables scalable ergonomic monitoring via automated REBA assessment, current deterministic systems suffer from hazardous overconfidence, providing definitive risk scores even when occlusion or lighting compromises data integrity. This study presents an information-theoretic framework that decomposes ...
Abstract: Construction workers face multifaceted physical and psychological health risks due to challenging and often unpredictable working environments. This study introduces a computational framework that uses artificial intelligence and natural language processing to extract and structure occupational health insights from 7,441 Reddit posts and comments related to construction. Focusing on eight ...
Abstract: As VR evacuation training and performance-based egress design increasingly rely on simulation, their credibility depends on empirical, space-specific motion benchmarks. However, VR evacuation research often lacks empirical, area-specific motion parameters for calibration. We demonstrate benchmark values from a full-scale multi-floor drill, reconstructed from multi-camera video as 2D trajectories in lobbies ...
Abstract: Facade shading systems can reduce cooling energy demand in office buildings located in tropical-humid climates, but the additional materials required may also increase embodied carbon emissions. Although many parametric facade optimization studies focus on operational energy and daylight performance, the temporal relationship between embodied carbon and operational carbon savings remains ...
Abstract: Building Information Modelling (BIM) is widely recognised as a key driver of digital transformation in the architecture, engineering and construction industry. However, its full potential remains underutilised, partly due to the vast and complex nature of BIM-related data. At the same time, Natural Language Processing (NLP), a rapidly evolving field ...
Abstract: In space-constrained modular construction, stochastic work content renders deterministic scheduling ineffective, often leading to production deadlocks. While Simulation-Based Optimization (SBO) addresses these dynamics, it typically lacks the operational utility for routine daily planning due to its high computational intensity. To bridge this gap, this study adopts a Design-Science Research (DSR) ...
Abstract: Industrialized construction relies on the tight integration of production, logistics, and on-site assembly, making the process highly sensitive to dynamic disturbances. To achieve rescheduling with minimum human intervention, this study proposes an intelligent rescheduling framework based on AI agents. First, we establish a unified Mixed-Integer Linear Programming (MILP) model to ...
Abstract: Indoor air quality management remains largely reactive due to limited multi-horizon forecasting capabilities and deployment challenges under real-world constraints. This paper presents a Digital Twin (DT)-enabled framework for multi-horizon indoor PM2.5 forecasting from 1 to 6 hours ahead using machine learning and multi-source environmental data from a university campus in ...
Abstract: Scaffolding accidents account for significant fatalities in the construction industry, with deviations between erected structures and approved design drawings being a leading cause. However, automated verification of scaffold compliance with design specifications remains an open challenge due to the difficulty of extracting geometric information from engineering drawings and matching them ...
Abstract: Urban façade deterioration, especially tile delamination and related surface damage, poses significant safety risks in dense urban environments. Conventional inspections based on manual visual assessment and acoustic sounding are labor-intensive, hazardous, and subjective. This paper presents an autonomous UAV-based façade inspection framework that combines oblique-photogrammetry-driven path planning with decision-level fusion ...
Abstract: This paper presents an AI-driven framework designed for automating and optimizing structural systems within a Building Information Modeling (BIM) workflow. The process begins by examining architectural layouts from IFC files to automatically define a valid search space for structural elements, ensuring that the solutions adhere to a set of structural ...
Abstract: Due to construction work's complexity and uncertainty, single simulation methods such as system dynamics (SD), discrete-event simulation (DES), and agent-based modelling (ABM) face significant limitations when applied independently. Hybrid simulation offers a promising solution, yet lacks both a comprehensive review and a structured framework for construction applications. This paper addresses ...
Abstract: While Virtual Reality (VR) provides a promising platform for construction safety training, generating virtual environments (VEs) that are tailored to specific project and task contexts remains a bottleneck. Current workflows, ranging from manual modeling to semi-automated BIM-to-VR pipelines, are often labor-intensive and result in static, idealized VEs that fail to ...
Abstract: Building Information Modeling (BIM) is widely used in construction management, particularly for mechanical, electrical, and plumbing (MEP) systems, where accurate pipeline information is critical for efficient operation and maintenance. In practice, however, as-built pipelines often deviate from the as-designed BIM models, and manually updating these models is time-consuming. This study ...
Abstract: Architectural design revision must satisfy user affective intent while preserving previously validated design qualities. We propose a BIM-integrated revision framework that (i) represents spatial experience as a multidimensional affective distribution, (ii) predicts the distribution from discrete design parameters using a lightweight transformer regressor, and (iii) selects a minimally disruptive revision ...
Abstract: As the safety of civil infrastructure is gaining more attention nowadays, inspection plays an important role in evaluating the condition of the structure. Ranking the level of defect is one of the most direct ways to understand the status of the building. Traditional inspections are based on handwritten records from ...
Abstract: Bridge point cloud instance segmentation is essential for digital twin construction and automated bridge management. However, deep learning-based approaches are limited by the scarcity of instance-level annotated real-world data and the domain gap between synthetic and real point clouds. In practical end-to-end Scan-to-Bridge Information Modeling(BrIM) pipelines, segmentation models must operate ...
Abstract: Construction projects generate heterogeneous data from schedules, design models, and site observations. Graph-based data models have recently been adopted to integrate these diverse data sources, but interacting with such models typically requires specialized technical knowledge, which limits their usability in construction practice. This study proposes a schema-aware Graph Retrieval-Augmented Generation ...
Abstract: Autonomous navigation in cluttered construction environments is challenging due to unstructured obstacles, GPS-denied conditions, and operational constraints on onboard computation. Optimization-based planners such as Covariant Hamiltonian Optimization for Motion Planning (CHOMP) offer obstacle-avoidance capabilities but are often unsuitable for high-frequency control due to their computational cost and sensitivity to local ...
Abstract: As digital fabrication systems in construction become more sensorized and controllable, they expose growing opportunities for data-driven process understanding that remain largely underutilized. Consequently, additive manufacturing workflows continue to rely on expert intuition and trial-and-error, which limits their industrial adoption in dynamic environments, such as construction sites. This paper presents ...
Abstract: Facilities management teams encounter complex HVAC maintenance scenarios as faults propagate across interconnected subsystems. This study presents HVACONNECT, a text-enriched heterogeneous graph learning framework for HVAC-related maintenance triage. Using six years of CMMS work orders and BAS signals from a 19-building campus, a heterogeneous graph capturing building hierarchies, HVAC systems, ...
Abstract: Recent studies have highlighted the necessity for a conceptual cost prediction model for Modular and Offsite Construction (MOC) projects. Barriers such as high costs and the absence of scientific methods for justifying cost comparisons between MOC and conventional construction methods impede the adoption of MOC. This study presents an ensemble ...
Abstract: Large language models (LLMs) evolve rapidly, yet their probabilistic generation is inherently non-traceable and leaves a persistent bottleneck in domain knowledge cognition. Retrieval augmentation via texts or knowledge-graphs (KGs) improves coverage and factual accuracy, but existing studies still depend largely on static knowledge structures and manually governed updating, so the ...
Abstract: This study explores the integration of Building Information Modeling (BIM)-based selective disassembly planning with robotic automation to facilitate the efficient and autonomous retrieval of building components, addressing critical challenges in the construction industry's transition toward a Circular Economy (CE). The BIM-based disassembly planning system developed in previous studies forms the ...
Abstract: Spraying processes, such as shotcrete, painting, and abrasive blasting, are critical to construction but remain labor-intensive, hazardous, and challenging to automate. Realistic and computationally efficient simulation models are crucial for supporting robotic process planning, training, and control. However, existing approaches are typically process-specific and lack generality across different material systems. ...
Abstract: Heritage buildings exhibit complex and irregular legacy environments that pose significant challenges for conservation. This study introduces a framework grounded on high-precision terrestrial laser scanning (TLS), which serves as input for both semantic segmentation of structural components and surface-level feature extraction. Leveraging these high-resolution TLS scans, the proposed approach supports ...
Abstract: The construction process requires close collaboration among multiple crews, which relies fundamentally on efficient task allocation. However, the performance and efficiency of existing scheduling approaches remain limited and are insufficient to support fine-grained decision-making. This paper proposes a scalable Reinforcement Learning (RL) framework for Multi-Crew Task Allocation (MCTA) in flexible ...
Abstract: Robotic bridge inspection systems are emerging as crucial technologies to improve efficiency and safety in infrastructure maintenance, particularly in accessibility-constrained environments. However, the challenge of processing multi-view data and integrating detection, quantification, and report generation into a unified workflow remains unresolved. To address this challenge, this paper proposes an integrated ...
Abstract: Inspection of concrete box-girders is essential for ensuring the structural performance of a PSM bridge throughout its service life. With advances in 3D sensing and point cloud processing, 3D data-based approaches have been increasingly applied to bridge inspection. This paper presents a 3D reconstruction pipeline for concrete box-girder interiors that ...
Abstract: Simulation-based optimization (SBO) for energy-efficient building design exemplifies a typical design-change process that remains disconnected from open BIM-enabled collaboration. The gap stems from IFC's limited capacity to represent the information generated by SBO. To overcome the limitations, this study proposes a novel IFC schema extension for LLM Agent-enhanced SBO, named ...
Abstract: Leakage detection in subway tunnel linings remain critical yet challenging tasks. Conventional deep learning models typically require tens of thousands of training samples that are scarce in engineering practice. Large Vision Models (LVMs) possess impressive general-purpose capabilities but fall short when applied to specialized downstream tasks. To address this gap, ...
Abstract: Conventional rebar quality inspection in construction relies heavily on manual measurements, which are time-consuming, error-prone, and impractical for dense reinforcement layouts. 3D Gaussian Splatting (3DGS) has recently emerged as a promising technique for vision-based automated rebar inspection, enabling dense 3D reconstruction under complex reinforcement configurations. Despite this potential, most existing ...
Abstract: The modular construction industry requires early and tightly coupled reasoning across spatial layout, structural systems, module interfaces, and building services under fabrication, transportation, and assembly constraints. With recent advances in generative artificial intelligence (AI) and the emergence of large language models (LLMs), there is growing interest in leveraging these technologies ...
Abstract: Autonomous mobile robots deployed in building environments require comprehensive spatial semantics for navigation and operation. However, while current Industry Foundation Classes (IFC) standards provide both geometric and semantic information, they lack robot-operational attributes. This paper presents a Point-of-Interest (POI) framework that enriches IFC building models with robot-oriented behavioral intelligence through ...
Abstract: Construction planning remains error-prone because existing automated approaches primarily assess final schedules and lack mechanisms to evaluate intermediate planning decisions. As a result, early-stage errors in task decomposition, sequencing, or duration estimation can propagate unnoticed across planning stages. This study proposes a process-aware framework for automated construction planning that embeds ...
Abstract: Building construction projects are socio-technical production systems in which variance, uncertainty and unpredictable events are the norm. Such systems need short-cycle Plan-Do-Check-Act (PDCA) loops to detect anomalies in production flow as early as possible so that remedial actions can be evaluated and well-informed decisions made. This is the goal of ...
Abstract: This study develops an automated disaster response report generation system that integrates large language models (LLMs), web scraping, and computer vision techniques. The system automatically retrieves data from 16 heterogeneous and distributed sources and, by combining Retrieval-Augmented Generation (RAG) with Chain-of-Thought (CoT) prompting, produces first drafts of disaster response reports ...
Abstract: Modular construction (MC) has attracted increasing attention as a promising strategy for carbon emission reduction in the context of global low-carbon transitions. However, accurate estimation and continuous tracking of lifecycle carbon emissions in MC projects remain limited, as existing studies predominantly focus on individual stages and provide fragmented evidence. To ...
Abstract: Construction productivity has remained stagnant for decades, underscoring the need for productivity improvement methods. Crew Balance Charts (CBCs) are an effective tool for analysing crew dynamics and identifying opportunities for productivity improvement. However, manual preparation of CBCs is time-consuming and limits scalability. This paper proposes a Computer Vision-based Crew Balance ...
Abstract: The Last Planner System® (LPS®) and Building Information Modeling (BIM) hold strong potential to improve construction productivity, yet their integration at the data-processing level remains limited. Information embedded in BIM models is often insufficiently structured and therefore not directly exploitable for LPS® implementation. This study proposes an approach that integrates ...
Abstract: Amidst escalating global climatic threats, effectively disseminating disaster preparedness knowledge to the public remains a critical challenge. This research introduces Disaster Knowledge King (DKK), an automated, dynamic quiz system designed to modernize disaster risk reduction (DRR) education. While traditional educational games suffer from static, "dead" databases that require manual updates, ...
Abstract: Construction projects frequently face disputes caused by unclear documentation, conflicting design versions, and accountability gaps among stakeholders. While Building Information Modeling (BIM) enhances collaboration, it currently lacks the immutable audit trails necessary for legal evidence. This paper explores the integration of blockchain technology with BIM to create transparent, tamper-proof systems ...
Abstract: The construction industry faces persistent economic, social, and environmental challenges. Construction robotics and lean construction have emerged as potential solutions, yet their integration remains underexplored. This paper introduces the notion of Robot-Oriented Lean Construction (ROLC) - a toolkit that adapts lean principles to the characteristics of construction robots, emphasizing synergy ...
Abstract: In high-context construction cultures like Japan, the Setup phase?comprising nemawashi (consensus-building) and dandori (operational preparation)?is critical for reliable execution. While BIM models are increasingly available, site engineers often lack the modelling expertise to evaluate them independently. This reliance on specialists creates bottlenecks, forcing planning back to manual quantity take-offs and ...
Abstract: Construction activities depend on multiple interrelated flows, materials, labor, equipment, workspace, and precedence, which introduce significant complexity into planning and control processes. Current simulation approaches typically focus on activity-level information while overlooking how these flows interact and evolve within the project environment. As a result, project managers lack effective tools ...
Abstract: Heavy rains and flooding due to global warming are becoming more frequent, increasing the risk of scouring and riverbed deformation. Low-cost and simple techniques can be effective for detecting these damages at an early stage, however, water currents, turbidity, and waves make surveys difficult. Therefore, we devised a method to ...
Abstract: Machine guidance using total stations (TSs) and global navigation satellite systems (GNSSs) is implemented to reduce labor and improve productivity in excavation work. However, a challenge with existing methods is that obstacles can interfere with guidance. Direct measurement of excavation surfaces using LiDAR can solve this problem, but it requires ...
Abstract: This paper introduces a graph-based method that derives robot-executable task structures from Building Information Models (BIM). While BIM models are widely employed to coordinate design and planning, they are seldom used to facilitate on-site automation. A crucial challenge is that they typically describe building elements and properties, but not the ...
Abstract: This paper presents a low-cost, high-precision laser-line sensor designed for construction-robotics applications such as on-site tiling, leveling, and tool localization. Unlike similar projects built around custom electronics and hardware, our approach uses a off-the-shelf microcontroller with camera module and a semi-transparent projection screen to measure the vertical position of a ...
Abstract: Scan-to-BIM procedures using terrestrial laser scanning (TLS) are accurate but require significant time and financial investment. This study evaluates a faster alternative utilizing mobile device-based LiDAR and a mobile application built on the Apple RoomPlan API. The research first assesses geometric reliability through point-cloud-to-BIM confidence indices, specifically Coverage, Distribution, and ...
Abstract: Digital Twin (DT) is an emerging technology that can be used for Indoor Environmental Quality (IEQ) optimization. However, its application in educational buildings, especially in the Indian context, is limited. Students spend a significant amount of their total time indoors in educational buildings, making effective IEQ management critical for their ...
Abstract: Bridge infrastructure is highly susceptible to deterioration due to continuous exposure to harsh environmental conditions and increasing traffic volumes and loads, necessitating frequent inspections to ensure structural safety and serviceability. Traditional bridge inspection approaches are often time-consuming, labor-intensive, error-prone, and disruptive to traffic operations, highlighting the need for more efficient, ...
Abstract: Service robots have been increasingly deployed across several infrastructure applications. ROS2 has become the widely used operating system for service robots due to its performance and compatibility. However, its default configuration lacks authentication, encryption, or access control. While prior studies have highlighted general ROS2 security weaknesses, the feasibility of reconnaissance ...
Abstract: Unobserved worker motions on active construction jobsites contribute to injuries, near-misses, and productivity loss; however, continuous monitoring remains difficult under real deployment constraints. This study presents a real-time worker activity monitoring framework that integrates lightweight Temporal Convolutional Networks (TCN) with wearable sensing and zone-level localization to enable continuous and ...
Abstract: A detailed understanding of the Urban Heat Island is crucial for energy efficiency, renewable integration, building system automation and demand response in district energy systems. Satellite images provide high resolution information about land surface temperatures (LST) at urban scale, but do not provide information about the air temperatures (AT), which ...
Abstract: Robotic facade installation across multiple floors necessitates millimeter-level geometric accuracy to ensure reliable anchor alignment and panel placement. The present paper proposes a survey-controlled sensing and modeling framework that integrates a construction tool with high-resolution close-range imaging to establish a consistent multi-floor reference frame. The utilization of facade anchors is ...
Abstract: On active jobsites, unobserved worker motions drive in- juries and near-misses, unsafe habits evade periodic audits, so reliable activity monitoring is essential to prevent delays and costly rework. In recent years, Inertial-Measurement- Unit(IMU) based motion recognition using Deep Learning approach is used widely in many ...
Abstract: This study provides a review of the literature regarding unmanned aerial vehicles (UAV) surveying for emergency facility construction, with a focus on data collection and processing methods. Synthesizing publications between 2016 and 2025, this study categorizes the applications of visible-light cameras, thermal infrared cameras, light detection and ranging (LiDAR), and ...
Abstract: Reconstructing internal structural elements remains a significant challenge in the construction industry due to limited access to concealed components. Muon tomography provides a non-destructive approach for capturing information about internal structural conditions, yet its application in construction remains limited, especially in 3D model reconstruction. This paper proposes a few-shot parameter ...
Abstract: Extrusion-based robotic additive manufacturing for construction-scale fabrication is highly sensitive to process faults such as unstable flow, clogging, and fluctuations in feed supply. Without continuous monitoring, these failures can cause prolonged non-deposition and geometric error. To address this limitation, this work presents a vision-guided supervisory framework that couples nozzle-centric process ...
Abstract: Safety remains a persistent challenge in the construction industry, where compliance with OSHA regulations forms the basis of most protective practices. Although these standards are comprehensive, they are distributed across lengthy and highly structured documents, making it difficult for practitioners to locate provisions relevant to specific tasks or hazards. This ...
Abstract: In recent years, the civil engineering and construction industry has faced a severe shortage of skilled workers due to a declining workforce and the aging of infrastructure. To address this challenge, we have been developing ROS2-TMS for Construction, a cyber-physical system (CPS) platform for earthwork sites that aims to automate ...
Abstract: Non-destructive Testing (NDT) is important for building structure inspection. Ground Penetrating Radar (GPR) is an effective non-destructive testing technique for building inspections, yet the complexity of the radargram data has limited its widespread adoption in engineering practice. This paper presents an approach for reconstructing raw GPR radargrams into 3D point ...
Abstract: Advances in robotic systems, sensing technologies, and artificial intelligence offer substantial potential to automate construction processes. However, current construction robotics applications are limited to specific tasks in isolated systems, lacking the integration required for coordinated, large-scale robotized construction that is applicable in a wide spectrum of trades. A key challenge ...
Abstract: Interoperability between industrial and construction digital models remains a persistent challenge, particularly in projects that require the integration of manufacturing machinery within building environments. Manufacturing workflows typically rely on CAD models based on the STEP standard, whereas the Architecture, Engineering, and Construction (AEC) sector adopts BIM methodologies founded on the ...
Abstract: In structural health monitoring (SHM) systems, sensor signals are often affected by multiple external load effects that are strongly coupled across different frequency bands, posing significant challenges for accurate signal interpretation. To address this issue, this paper proposes a two-stage blind source separation method aimed at effectively decoupling sensor signals ...
Abstract: Construction sites are complex and dynamic work environments with frequent changes in layout due to material handling and a diverse workforce presence, necessitating adaptable safety measures. Falls from height remain the primary cause of fatalities in construction, particularly impacting small firms with limited resources. Conventional manual construction site layout inspections ...
Abstract: Identifying construction equipment activity states is essential for understanding and optimizing earthmoving operations. Trajectory data can be used to infer semantic equipment activities, while it often leads to inconsistent results when simple rule-based methods are applied. This paper presents a trajectory-based activity recognition approach using finite-state machines (FSMs) for earthmoving ...
Abstract: In many urban areas, under-bridge spaces have become common gathering spots for older adults, offering venues for leisure, social interaction, and physical activity. Various policies and community initiatives have been introduced to improve the quality and usability of these spaces. However, efficiently and automatically assessing the elder-friendliness of such environments ...
Abstract: Construction and demolition waste (CDW) sorting is increasingly being automated using robotics and AI to improve safety, efficiency, and material recovery. However, developing and validating perception and control algorithms directly on physical conveyor systems is costly, time consuming, and difficult to reproduce due to variability in object layouts and operating ...
Abstract: Building Information Modeling (BIM) has become a central platform for digital building representation; however, the creation and manipulation of BIM models remain largely manual, and accessible primarily to domain experts. At the same time, architectural information is still commonly produced in 2D drawing formats, creating inefficiencies in translating design intent ...
Abstract: Precise geolocation of social media is vital for real-time urban flood monitoring. However, researchers are frequently hindered by the scarcity of geotags due to privacy protections and the limitations of coarse localization based on posting frequency, which fails to pinpoint street-level flood sites. We present a robust end-to-end framework that ...
Abstract: Productivity and disruption monitoring remains a persistent challenge in construction projects. Although digital technologies such as Internet of Things (IoT) sensors, Building Information Modeling (BIM), computer vision, and AI-driven analytics have been introduced, practitioners often lack systematic guidance on aligning these tools with specific monitoring objectives to improve productivity and ...
Abstract: Infrared thermography is a widely adopted non-destructive technique for infrastructure inspection, enabling the detection of defects such as thermal bridges, insulation deficiencies, moisture accumulation, and material degradation. In recent years, the use of this technique alongside Unmanned Aerial Vehicles (UAVs) has attracted growing interest due to their ability to access ...
Abstract: Modular Construction Manufacturing (MCM) has attracted considerable attention in recent years due to its potential to increase productivity, minimize waste, and improve the predictability of building projects. Despite these benefits, effective coordination and real-time information exchange among multiple stakeholders in the delivery of modular construction remains a significant challenge, particularly ...
Abstract: Accurate and real-time recognition of construction workers' behaviors is critical to intelligent safety supervision and productivity assessment. However, existing skeleton-based methods struggle to balance recognition accuracy with inference latency, while lacking robustness against noisy keypoints caused by frequent occlusions. This study proposes RepCA-GCN, a high-efficiency architecture specifically tailored for construction ...
Abstract: The current prefabricated material tracking method in construction projects usually uses RFID or GPS technology. This technique requires tagging materials, which can be labour-intensive for large volumes of materials on a construction site. Deep learning-based object detection can provide a solution for material tracking in construction projects without the need ...
Abstract: The construction industry is characterized by a high degree of structural complexity, making it difficult to achieve a high level of modularization at the current stage. During operations involving robotic arms, human operators are often required to enter the robot's workspace at certain stages to perform fine adjustments, which may ...
Abstract: New Zealand's school building programme is deploying modern methods of construction (MMC) and nationwide IoT sensors faster than almost any other public sector, yet this descriptive multi-case analysis of 17 schools shows that this is disconnected from net-zero goals. Government data indicate that 60 % of new classrooms now use ...
Abstract: Smart community development is pivotal for modernizing grassroots governance. However, few studies have established a systematic classification model for residents' multi-dimensional needs. To fill this research gap, this paper constructs a comprehensive demand analysis framework covering three core dimensions: community safety, livability services and community governance. Based on the survey ...
Abstract: Worksite trailers are widely used as rapidly deployable prefabricated units, yet their design process often struggles to balance standardization for efficiency with customization for diverse usage needs. This paper presents an interactive early-stage design workflow that integrates Large Language Models (LLMs) with Building Information Modeling (BIM) to accelerate requirement interpretation ...
Abstract: Automated productivity monitoring and robotic learning in prefabrication are hindered by the scarcity of structured demonstration data. To address this, this study investigates a Vision-Language Model (VLM) workflow for extracting engineering logs from wall panel assembly videos. We conducted a comparative experiment using a 2 × 3 factorial design to ...
Abstract: Mechanical, Electrical, and Plumbing (MEP) engineering drawings serve as the contractual foundation for construction projects, yet verifying their consistency remains manual and error-prone. While Vision-Language Models show promise, they exhibit symbol fragility, spatial reasoning failures, and limited cross-document reasoning on technical drawings. We propose SAGE, a multi-agent framework employing: (1) ...
Abstract: Construction project operations remain difficult to automate due to dynamic site conditions, limited workflow standardization, and the absence of adaptable robotic platforms. General-purpose service robots (GPSRs), while mature in navigation, sensing, and data acquisition, lack a structured methodology for deployment in construction environments. This paper proposes a domain-specific systems engineering ...
Abstract: The construction industry is undergoing rapid transformation driven by digitalization, automation, and changing workforce structures. Despite these shifts, career guidance practices within the sector remain largely manual, fragmented, and poorly aligned with dynamic labor market conditions, limiting their effectiveness for workforce planning and innovation adoption. This study addresses this gap ...
Abstract: This paper examines how Construction 4.0 technologies tackle the challenges of site-based women in construction. The study also examines the role of Construction 4.0 in offering a competitive advantage to women in these site-based roles. The study adopts a qualitative method using semi-structured interviews with 29 women, of whom nine ...
Abstract: Building Information Modelling (BIM) has emerged as a transformative digital technology that can improve productivity, reduce errors, and enhance collaboration across construction projects worldwide. Despite its well-recognised benefits, the Architecture, Engineering, and Construction (AEC) industry in Iran continues to encounter substantial challenges in BIM adoption. Research exploring effective solutions tailored ...
Abstract: Despite rapid technological innovation, the construction industry's slow adoption highlights a critical lack of systematic, validated frameworks for evaluating emerging methods against established practices. This study presents the development of a conceptual decision-support framework for assessing steel connection systems. It compares an innovative intermeshed steel connection (ISC) with the existing ...
Abstract: This paper presents a systematic review of intelligent construction robots and their readiness for real-world deployment. Intelligent construction robots are defined as software-based or physically embodied agents that operate in a Perception-Reasoning-Action loop?sensing the environment, reasoning over observations and/or human commands, and executing actions through control or actuation. Using a ...
Abstract: The AECO sector increasingly calls for digital transformation to support sustainable development and a more human-centred built environment. In this context, the Society 5.0 dimensions, which emphasise human-centred values, cyber-physical intelligence, socio-technical embeddedness, societal challenge orientation and system-level impact potential, provide a lens for examining how digital technologies are ...
Abstract: This study examines the emerging role of Artificial Intelligence (AI) and blockchain in construction supply chain research through bibliometric and thematic analysis of the existing literature. Using VOSviewer, the study identifies key research trends, thematic clusters, and knowledge structures related to supply chain transparency, automation, optimisation, and adoption barriers. The ...
Abstract: Factories are a key part of industrial value creation and have a significant impact on the environment throughout the entire supply and production chain. Therefore, a sustainability assessment of both the building and the production system requires a holistic, interdisciplinary perspective on the factory system, which cannot be achieved using ...
Abstract: We created a support system for inspecting the exterior appearance by loading a three-dimensional photogrammetric model into a virtual space. This system shows a photograph of what the observer sees in virtual space. To realize this system, we devised a method to represent buildings as voxels, check whether the ...
Abstract: The resilience of construction supply chains has become a critical issue in dense urban environments, particularly for modular construction systems that rely on large and time-sensitive deliveries. While existing studies have extensively addressed construction logistics and supply-chain resilience at strategic and regional levels, the last-mile delivery phase within urban areas ...
Abstract: Underground utility information is frequently incomplete, inconsistent, or of uncertain confidence, leading to safety risks, rework, and inefficiencies during construction and maintenance activities. Although existing standards, sensing technologies, and digital platforms define data quality concepts and investigation methods, their application within construction workflows remains fragmented and largely project-specific. In particular, ...
Abstract: The rapid change in global climatic conditions and the increasing number of severe climate events have urged the demand for rapid, installable, customized, and sustainable housing solutions for climate-affected and displaced populations. To address this concern, this study develops a modular housing system tailored for climate displacements that integrates prefabricated ...