This paper proposes the Ray-Based Laser Scanning and Intersection Algorithm (RBLSIA) to generate synthetic point clouds for Mechanical, Electrical, and Plumbing (MEP) systems using BIM models, addressing the lack of MEP datasets for deep learning-based semantic segmentation. Twenty comparative experiments were conducted to assess the performance across different training datasets, synthetic point cloud generation methods.The results show that RBLSIA-generated synthetic point clouds outperform those from uniform sampling by 3.32% in mean Intersection over Union (mIoU). Additionally, increasing the volume of synthetic samples improves overall accuracy (OA) and mIoU, surpassing the performance of models trained with real point clouds.
The railway sector struggles with infrastructure inefficiencies due to traditional maintenance methods, resulting in high costs and unplanned disruptions. This paper provides a conceptual framework that integrates Digital Twins (DT) and Cyber-Physical Systems (CPS) to enhance predictive maintenance (PdM). By establishing a technical architecture for real-time monitoring and fault detection, the framework will facilitate seamless data exchange for automated decision-making. The findings contribute to advancing intelligent railway maintenance, fostering sustainability, and enhancing resilience in railway operations. This research provides actionable insights for industry stakeholders, supporting the transition towards data-driven, adaptive maintenance strategies in modern rail networks.
In the era of mass customization, industrialized construction could use product configurators to balance standardization and design flexibility, therefore maintaining productivity and economic efficiency. Previous studies have proposed conceptual frameworks for product configurators, but few studies illustrate how they could be implemented in a practical building project. This paper extends the framework proposed by the author in a previous study to prototype development and explains the implementation process in collaboration with a construction company. The study contributes to a more generalized framework enabled by graph-based configuration rules and a scalable implementation plan of configurators in practice.
This paper advocates a holistic socio-techno-economic approach to IT system-building for Life Cycle Asset Information Management (LCAIM) in the built environment, challenging the current technology-focused research trajectory. Drawing precedence from paradigm transitions in other related sectors, a Design Science Research (DSR) methodology is used to propose and prototype a framework as a software UI, evaluating it using expert focus groups. The paper contributes a framework that aligns IT system design with the environmental policy landscape, emphasising considerations beyond technical aspects, such as trust and scalability. The results demonstrate the frameworks effectiveness in addressing the broader requirements of future LCAIM systems.
Real estate renovation processes face challenges of fragmentation, limited transparency, and poor communication. This study presents DesignHome VR, a patent-pending platform integrating immersive reality with workflow management to address these issues. The cloud-based system combines spatial data processing with collaborative interfaces, demonstrating improved project efficiency through reduced review cycles and enhanced cost accuracy. A case study validates the platform's effectiveness in design accuracy and workflow simplification. The innovation lies in its cohesive framework that simultaneously addresses visualization needs and stakeholder collaboration, democratizing renovation processes through immersive technology.
The construction industry is evolving with BIM, AI, VR, and robotics, yet traditional methods persist, with paper-based drawings and legacy software still widely used. Meanwhile, construction education is shifting towards advanced technologies, presenting challenges in balancing modern innovations with traditional skills. This study surveys 22 general contractors primarily working in the Mississippi area to assess information delivery methods and hardware and software used. Results reveal widespread reliance on paper-based and CAD files, and low utilization of advanced technologies. The findings inform recommendations for updating construction management curricula to balance cutting-edge and traditional approaches, ensuring graduates are industry-ready.
We present Shiva, a workflow-integrated, robotic 3Dcapture tool for 3Dscanning small-to-medium sized objects. Addressing challenges in speed, flexibility, and automation. Shiva integrates 3Dscanning with robotic position, enabling a precise and rapid point cloud generation with automatic alignment with support for single and dual robot configurations. Implemented as a Grasshopper plugin, it can be integrated in computational workflows in need of feedback loops with physical objects, such as in fabrication with conformal or adaptive 3Dprinting. The developed tool is validated in use cases with bio-based materials to automate the quality of digital inspection, adaptive reuse, and real-time feedback through projection mapping.
Artificial Intelligence (AI) is revolutionizing the Architecture, Engineering, and Construction (AEC) industry. This paper reports on an elective course, "Advanced Exploration of AI Use Cases in AEC," where student groups of 2-3 examined AI’s potential through a structured approach. Students selected a broad topic, conducted traditional literature research, and used AI tools to refine their focus on specific AEC challenges. They evaluated both the research process and the capabilities of AI tools, models, and methods in addressing AEC applications. The course culminated in presentations comparing traditional and AI-based approaches, highlighting AI’s impact on research efficiency and practical problem-solving.
Project schedules are critical for construction management, yet they remain inaccessible to many professionals due to complex scheduling software and costly licenses. As a result, site engineers and last planners rely on static, outdated schedule copies, limiting proactive decision-making. This study proposes an AI-powered schedule management application that enables natural language interaction with project schedules. By leveraging knowledge graphs (KG) and NLP, the app enhances accessibility, reduces inefficiencies, and ensures real-time interaction with updated schedules. The results demonstrate improved usability and engagement with schedules. This research contributes an AI-powered schedule management solution to bridge the gap between scheduling expertise and project execution.
Few attempts have targeted leveling resources in linearschedules by mainly changing production rates. However, thisled, in some instances, to a decrease in productivity. Therefore,the paper presents a two-phase model that: (1) generates thelinear schedule and calculates start and end dates and availablefree float days for each activity, and (2) performs resourceleveling according to a newly-designed parameter called“average deviation of data points from their mean”. The outputis a leveled production rate-conserved linear schedule, whichclearly highlights the study’s contribution in providing aresource leveling tool for linear schedules that preserves theproduction rates.
