Search results

The practice of artificial intelligence (AI) is increasingly being promoted by technology developers. However, its adoption rate is still reported as low in the construction industry due to a lack of expertise and the limited reliable applications for AI technology. Hence, this paper aims to present the detailed outcome of experimentations evaluating the applicability and the performance of AI object detection algorithms for construction modular object detection.

This paper provides a thorough evaluation of two deep learning algorithms for object detection, including the faster region-based convolutional neural network (faster RCNN) and single shot multi-box detector (SSD). Two types of metrics are also presented; first, the average recall and mean average precision by image pixels; second, the recall and precision by counting. To conduct the experiments using the selected algorithms, four infrastructure and building construction sites are chosen to collect the required data, including a total of 990 images of three different but common modular objects, including modular panels, safety barricades and site fences.

The results of the comprehensive evaluation of the algorithms show that the performance of faster RCNN and SSD depends on the context that detection occurs. Indeed, surrounding objects and the backgrounds of the objects affect the level of accuracy obtained from the AI analysis and may particularly effect precision and recall. The analysis of loss lines shows that the loss lines for selected objects depend on both their geometry and the image background. The results on selected objects show that faster RCNN offers higher accuracy than SSD for detection of selected objects.

The results show that modular object detection is crucial in construction for the achievement of the required information for project quality and safety objectives. The detection process can significantly improve monitoring object installation progress in an accurate and machine-based manner avoiding human errors. The results of this paper are limited to three construction sites, but future investigations can cover more tasks or objects from different construction sites in a fully automated manner.

This paper’s originality lies in offering new AI applications in modular construction, using a large first-hand data set collected from three construction sites. Furthermore, the paper presents the scientific evaluation results of implementing recent object detection algorithms across a set of extended metrics using the original training and validation data sets to improve the generalisability of the experimentation. This paper also provides the practitioners and scholars with a workflow on AI applications in the modular context and the first-hand referencing data.

The purpose of this paper is to analyze friction and controversies with interaction processes and their effects on forming new resource interfaces, through the lens of boundary objects.

The empirical setting consists of two organizations that are trying to enhance their competitive advantage through digitalization. During the process of data collection four different boundary objects were identified. The study illustrates how these boundary objects were characterized in terms of their modularity, standardization, abstractness and tangibility. This paper provides an analysis of how respondents perceived that the development of these boundary objects affected the creation of novel resource interfaces, and the resulting friction and controversy between new and old structures.

The study concludes that within a producer–user setting a focal boundary object will take on tangible and standardized properties, and the interaction process will expose friction in terms of both power struggles and resource incompatibilities. On the other hand, a boundary object’s modularity gives the actors central to the interaction room to maneuver and avoid resource incompatibilities and the development setting will hence be characterized by controversies.

The analysis indicates that the way individuals perceive boundary objects is central to interaction processes, answering calls for studies that investigate the role of objects within subject-to-object interaction.

This paper aims to propose a new deep learning technique to detect the type of material to improve automated construction quality monitoring.

A new data augmentation approach that has improved the model robustness against different illumination conditions and overfitting is proposed. This study uses data augmentation at test time and adds outlier samples to training set to prevent over-fitted network training. For data augmentation at test time, five segments are extracted from each sample image and fed to the network. For these images, the network outputting average values is used as the final prediction. Then, the proposed approach is evaluated on multiple deep networks used as material classifiers. The fully connected layers are removed from the end of the networks, and only convolutional layers are retained.

The proposed method is evaluated on recognizing 11 types of building materials which include 1,231 images taken from several construction sites. Each image resolution is 4,000 × 3,000. The images are captured with different illumination and camera positions. Different illumination conditions lead to trained networks that are more robust against various environmental conditions. Using VGG16 model, an accuracy of 97.35% is achieved outperforming existing approaches.

It is believed that the proposed method presents a new and robust tool for detecting and classifying different material types. The automated detection of material will aid to monitor the quality and see whether the right type of material has been used in the project based on contract specifications. In addition, the proposed model can be used as a guideline for performing quality control (QC) in construction projects based on project quality plan. It can also be used as an input for automated progress monitoring because the material type detection will provide a critical input for object detection.

Several studies have been conducted to perform quality management, but there are some issues that need to be addressed. In most previous studies, a very limited number of material types were examined. In addition, although some studies have reported high accuracy to detect material types (Bunrit et al., 2020), their accuracy is dramatically reduced when they are used to detect materials with similar texture and color. In this research, the authors propose a new method to solve the mentioned shortcomings.

In the construction industry, building facade systems are gaining significance as reflected in the number of facade product innovations available on the market. While research studies on the adoption of these facade product innovations are comparatively limited in number. An awareness of the advantages and impacts that a product innovation brings is critical in its adoption, and therefore the purpose of this paper is to develop a classification framework for the use of producers/vendors present the advantages and impacts of their innovative facade products to the architects.

The study, in addition to literature review, consisted of four main steps without distinct boundaries, namely, (1) case study sample selection regarding facade products, (2) data collection through review of published documents, analysis of the data via content analysis and generation of a preliminary framework for the evaluation of facade product innovations, (3) further data collection through semi-structured interviews and analysis of the products and (4) redefining innovation types and finalizing classification framework through assessment of new facade products.

Innovation types may vary according to the product groups under investigation and viewpoints of the classifications. Selected facade products were evaluated in terms of their specific characteristics through identified viewpoints: as a physical object and as an agent. It was revealed that assessing a product innovation in terms of changes in its structural composition and impact on other related physical objects/processes can change the type of innovation. Additionally, insights about improved characteristics specific to facade products, forms of change in these products and architects' new actions in the facade design process were obtained.

Only the products stated by the producers/vendors as being new to the Turkish market at the time they were introduced were evaluated in terms of all aspects proposed in the framework. For these products and others that were evaluated, the product data obtained from the producers/vendors have been deemed to be accurate. Research on equivalent products in the market has not been conducted.

Studies identifying innovation types in the facade industry are very rare. The proposed framework can be regarded as a detailed investigation of product innovation types in the facade industry which offers component-system level improvement/change analysis. Moreover, it can be an effective tool for producers/vendors to ensure that architects have knowledge about their innovative products and their impacts on facade design and also to facilitate the adoption of these products.

This study aims to introduce the DoraHand, and the basic capability and performance have been verified in this paper. Besides the idea of sharing modular design and sensor design, the authors want to deliver an affordable and practical dexterous hand to the research area to contribute to the robotic manipulation area.

This paper introduced the DoraHand, a novel scalable and practical modular dexterous hand, which, adopting modular finger and palm design, fully actuated joint and tactile sensors, can improve the dexterity for robotic manipulation and lower the complexity of maintenance. A series of experiments are delivered to verify the performance of the hand and sensor module.

The parameters of the DoraHand are verified and suitable for the research of robotics manipulation area, the sensing capability has been tested with the static experiment and the slip prediction algorithm. And, the advantage of modular design and extensible interface have been verified by the real application.

The authors continue improving the DoraHand and extend it to more different applications. The authors want to make the DoraHand as a basic research platform in the robotic manipulation area.

The DoraHand has been sent to more than ten different research institutes for different research applications. The authors continue working on this hand for better performance, easier usage and more affordability.

This kind of dexterous hand can help researchers get rid of complex physical issues and pay more attention to the algorithm part; it can help to make robotic manipulation work more popular.

The key design in the DoraHand is the modular finger and sensing module. With the special design in mechanical and electrical parts, the authors build reliable hardware and can support the diversity requirement in the robotic manipulation area. The hand with tactile sensing capability can be used in more research and applications with its extensibility.

Rather than organize as traditional firms, many of today’s companies organize as platforms that sit at the nexus of multiple exchange and production relationships. This chapter considers a most basic question of organization in platform contexts: the choice of boundaries. Herein, I investigate how classical economic theories of firm boundaries apply to platform-based organization and empirically study how executives made boundary choices in response to changing market and technical challenges in the early mobile computing industry (the predecessor to today’s smartphones). Rather than a strict or unavoidable tradeoff between “openness-versus-control,” most successful platform owners chose their boundaries in a way to simultaneously open-up to outside developers while maintaining coordination across the entire system.

This paper focuses on the development of a distributed rapid prototyping system via the Internet to form a framework of Internet prototyping and manufacturing for the support of effective product development. The proposed methodology is targeted at a wide audience using a disparate range of computer systems to access remotely located rapid prototyping facilities via the Internet for prototype fabrication. The methodology is useful for both educational research for teaching evolving rapid prototyping technologies and remote scientific visualization. This approach is based on the merger of object‐oriented modular software architecture and client server communications for the remote control of rapid prototyping hardware (called fused deposition modeling) via the Internet. Other Web tools are used to allow the remote user to have higher interactivity with the server applications that have a direct link with the front‐end terminals controlling the rapid prototyping hardware.

This paper aims to present an approach for the simulation of a heterogeneous robotic cell. The simulation enables the cell’s developers to conveniently compare the performance of alternative cell configurations. The approach combines the use of multiple available simulation tools, with a custom holonic cell controller. This overcomes the limitation of currently available robot simulation packages by allowing integration of multiple simulation tools including multiple vendor simulation packages.

A feeding cell was developed as a case study representing a typical robotic application. The case study would compare two configurations of the cell, namely, eye-in-hand vision and fixed-camera vision. The authors developed the physical cell in parallel with the simulated cell to validate its performance. Then they used simulation to scale the models (by adding subsystems) and shortlist suitable cell configurations based on initial capital investment and throughput rate per unit cost. The feeding cell consisted of a six-degree of freedom industrial robot (KUKA KR16), two smart cameras (Cognex ism-1100 and DVT Legend 500), an industrial PC (Beckhoff) and custom reconfigurable singulation units.

The approach presented here allows the combination of dissimilar simulation models constructed for the above mentioned case study. Experiments showed the model developed in this approach could reasonably predict various eye-in-hand and fixed-camera systems’ performance. Combining the holonic controller with the simulation allows developers to easily compare the performance of a variety of configurations. The use of a common communication platform allowed the communication between multiple simulation packages, allowing multi-vendor simulation, thereby overcoming current limitation in simulation software.

The case study developed here is considered a typical feeding and assembly application. This is however very different from other robotic applications which should be explored in separate case studies. Simulation packages with the same communication interface as the physical resource can be integrated. If the communication interface is not available, other means of simulation can be used. The case study findings are limited to the specific products being used and their simulation packages. However, these are indicative of typical industry technologies available. Only real-time simulations were considered.

This simulation-based approach allows designers to quickly quantify the performance of alternative system configurations (eye-in-hand or fixed camera in this case) and scale, thereby enabling them to better optimize robotic cell designs. In addition, the holonic control system’s modular control interface allows for the development of the higher-level controller without hardware and easy replacement of the lower level components with other hardware or simulation models.

The combination of a holonic control system with a simulation to replace hardware is shown to be a useful tool. The inherent modularity of holonic control systems allows that multiple simulation components be connected, thereby overcoming the limitation of vendor-specific simulation packages.

The purpose of this paper is to guide companies in conducting benchmarking studies of their manufacturing processes by viewing across industries, locations and products. In particular, the proposed framework can help corporate decision makers in terms of production footprint and site location studies. The level of benchmarking performance can be measured by evaluating defined benchmarking evaluation profiles.

This paper develops a tool to operationalize value-added manufacturing processes for benchmarking evaluations. In this context, an object-oriented database structure has been developed for the business areas such as product development, manufacturing and assembly. This paper focuses on manufacturing processes. Furthermore, a framework for applying high-performance benchmarking has been developed and applied in a case study.

This paper shows that object class-oriented modeling approach can be applied to manufacturing processes. The higher the degree of independence in terms of locations, industry sectors and products, the more powerful thus a higher performance of benchmarking is achieved. The performance level of benchmarking has been defined by proving and demonstrating higher and lower performance levels. The high-performance benchmarking tool has been successfully applied to a production footprint case study.

This paper takes up the superiority of process benchmarking that has been the focus of numerous research papers on benchmarking techniques in the past. The potential of process benchmarking has been enhanced and operationalized as a tool. A classification logic for benchmarking evaluation profiles has been developed and integrated in the overall tool set. The model helps decision makers to configure their benchmarking studies tailored to their strategic entrepreneurial questions and to guide them to achieve a higher benchmarking performance level.