Search

Learning building information modeling (BIM) systems has always been a challenge for BIM adoption. Although groundbreaking performances of large language models (LLMs) have inspired many researchers to consider an LLM as a potential BIM control method using natural language, a specific method of utilizing LLMs for automated BIM model detailing has not yet been proposed. This paper proposes an LLM-BIM chaining framework to enable architectural design detailing using natural language, instead of using menu-based user interfaces, named "Natural-language-based Architectural Detailing through Interaction with AI (NADIA)". The NADIA framework is based on three main approaches: 1) separating the specification of the wall layers from the creation of the wall layers; 2) appropriate instruction prompting to guide the LLM to minimize irrational responses and produce engineering rational details; and 3) LLM-BIM chaining to seamlessly link a BIM authoring tool and an LLM. The effectiveness of NADIA was validated based on two main aspects: its accuracy in generating details that adhere to specified design requirements from users—as a design assistant—and its compliance with general engineering requirements—as a design consultant. The validation was achieved through tasks that involved generating 240 and 1,920 exterior wall details, respectively. NADIA achieved an average accuracy of 83.33% in generating logically coherent details in line with the required design conditions. For thermal performance requirements, it demonstrated a mean accuracy of 98.54% in complying with the American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) 20.1–2019 standard. Despite being in its early stages, NADIA's potential for developing and refining architectural details through natural language-based interactions between architects and machines is promising. [ABSTRACT FROM AUTHOR]

Nature has undergone millions of years of evolution, enabling organisms to adapt and survive in their environment. The remarkable features developed by these organisms serve as a rich source of inspiration for designers involved in engineering design. However, the effective application of bio-inspired design faces challenges due to the gap between biology and engineering, as well as the limited level of design automation. This paper proposes a bio-inspired generative design framework (BIGD), containing three main steps of dataset building, generator modeling, and design evaluation, which aims to automatically produce innovative designs by synthesizing a diverse range of natural designs using deep generative models. Specifically, a computational workflow is established for automated bio-inspired wing shape synthesis. The process of constructing a dataset typically involves web crawling data from various sources, followed by data preprocessing to ensure clarity. The data is then structured with prior knowledge from the biological domain and outliers are excluded. Finally, design knowledge is extracted through data mining and knowledge extraction techniques. In the case of a flapping wing shape design, the deep generative model is constructed using a bird wing dataset created with a strong foundation in biological domain knowledge. The wings generated by BIGD exhibit superior lift performance across a range of working conditions, showcasing the advantages of using BIGD to directionally guide the evolution of generative models based on biological knowledge. This research highlights the potential of using computational methods in bio-inspired design to rapidly generate innovative and high-performance designs. [ABSTRACT FROM AUTHOR]

Abstract Purpose The purpose of this research was to develop a case-based method for analyzing online customer reviews and extracting customer preferences through an integration of text mining and Kansei Engineering (KE) in an effort to achieve conceptual data-driven design automation and to successfully identify future trends in a particular consumer product. Design/Methodology/Approach This study's model merges text mining and KE to extract key descriptive Kansei terminology according to actual customer reviews and use it to forecast consumer preferred product design while reducing certain repetitious tasks of designers. This work first collects online product reviews using text mining. Then, through the application of KE, the customer-preferred design components are identified and incorporated into the product design specifications. Finally, an Application Programming Interface (API) is developed to automatically generate a CAD preliminary design. Case Study A road bike case study is provided to demonstrate the practicality of proposed method. The online reviews are collected from Amazon.com. The related design elements are classified into six key components which can be modified in the proposed conceptual design automation system. Originality/value This is the first paper that has applied text mining and KE for use in product development. This work can also reduce the time and cost of product design through the automation of repetitive design tasks. The conceptual design automation system is valuable for designers wishing to identify customer needs and to generate engineering drawings in a timely manner without significant repetition during the design process. [ABSTRACT FROM AUTHOR]

• A new method of parametrically and automatically constructing fractal models or complex geometric objects with self-similarity features in a CAD system is proposed. • A new type of feature, named CAD fractal feature is proposed to bridge the gap between CAD and fractal geometry. • The concept of iterative construction of a CAD model is presented and multilevel representation of a iteratively constructed CAD model with different geometric complexities is realized. • The design intent is captured and knowledge for building associative relationships among geometry entities are encapsulated in a dynamic geometric generator with extracted variable design parameters. • An extruded Koch Snowflake CAD model with parametric geometric variants is constructed automatically. Fractal geometry has been widely applied in computer graphics to visually represent natural objects, which cannot be easily represented by Euclidean geometry. Other than visual representation, parametric fractal-like CAD models have been used in industry for the design and manufacture of products with self-similarity geometric features. However, contemporary CAD systems cannot construct CAD fractal models efficiently, as most CAD fractal models are still built manually by designers. The construction and modifications for such CAD fractal models are tedious and time-consuming. This paper aims to build and modify CAD fractal models more efficiently in CAD systems. A method for automatically modeling parametric fractal-like geometry objects in CAD systems is proposed. Feature-based modeling methods have been applied for realizing algorithm-based CAD fractal model generation. A new type of feature, called CAD fractal feature is proposed and implemented to bridge the gap between CAD and fractal geometry. The fractal-like CAD model is constructed iteratively with a dynamic generator for each iteration cycle. Design parameters are extracted as inputs for dynamic fractal generators, which allows users to define, parameterize, generate and edit a fractal design with variable parameters. The parameters of the fractal generator can also take different values for different model construction iterations, which makes the generated models not necessarily to be strictly fractal but can also be fractal-like. The iterative generation method (IGM) is proposed as a built-in method in a CAD fractal feature for modeling complex but self-similar geometric models in an objected-oriented approach with high efficiency and sufficient design flexibility. As a result, designers do not need advanced knowledge of fractal geometry to build and modify CAD models with self-similar features. An extruded Koch Snowflake CAD model with parametric geometric variants has been built as a case study for demonstration of the proposed modeling method. [ABSTRACT FROM AUTHOR]

Knowledge-based engineering (KBE) is a research field, which evolved in the 1980s within the context of artificial intelligence (AI), computer aided design (CAD) and computer programming. KBE is often referred to methods, processes and tools for knowledge reuse and design automation promising time and cost benefits in engineering design and originally especially in aerospace and automotive industry. Despite the long period of existence, theoretical foundations of KBE are weak and scientific literature is quite dispersed. About a decade ago, a review by Verhagen et al. collected the state of research and identified future research challenges, which has gained impact within the KBE research community. Therefore, within our review we aim at analysing the evolution of those research challenges in KBE literature during the last decade and further provide an overview of current shortcomings from a design research perspective. The analysis comprises 160 publications from the period of 2012 to 2021. Key findings from the review show, that the challenges from 2012 remain unsolved. We further derived that reuse and adoption of KBE approaches from literature is impeded by (1) a missing conceptualization of KBE research, (2) insufficient links from knowledge source to automation and (3) a tendency of insufficient distinction of the application area. [ABSTRACT FROM AUTHOR]

Generative design provides a promising algorithmic solution for mass customization of products, improving both product variety and design efficiency. However, the current designer-driven generative design formulates the automated program in a manual manner and has insufficient ability to satisfy the diverse needs of individuals. In this work, we propose a data-driven generative design framework by integrating multiple types of data to improve the automation level and performance of detail design to boost design efficiency and improve user satisfaction. A computational workflow including automated shape synthesis and structure design methods is established. More specifically, existing designs selected based on user preferences are utilized in the shape synthesis for creating generative models. For structural design, user-product interaction data gathered by sensors are used as inputs for controlling the spatial distributions of heterogeneous lattice structures. Finally, the proposed concept and workflow are demonstrated with a bike saddle design with a personalized shape and inner structures to be manufactured with additive manufacturing. [ABSTRACT FROM AUTHOR]

Reinforcement Learning has proven to be capable of solving complex tasks like playing video games, robotics control, speech or image recognition and processing. Transferring Reinforcement Learning into engineering design helps to overcome two current issues of data-driven Design Automation in engineering design. First, dealing with sparse training data resulting from differing design samples. Second, overcoming the limited number of samples in the training data as consequence of short or insufficient product history. To introduce an alternative approach for Design Automation, this contribution studies feasibility, training effort and transferability of Reinforcement Learning in engineering design. The presented method maps engineering requirements and parametric models into learning environments and provides a novel approach for design automation. In addition to that, the contribution summarises the hyperparameters, which design engineers have to set prior to training, and introduces a novel transfer learning concept for Reinforcement Learning in related design tasks. The support is probed by design tasks of performance-oriented bike parts. Case-independent indicators are presented to estimate the case-specific training effort, the effects of hyperparameter variation and the effects of transferring a pretrained agent to related design tasks. Finally, the findings are used to compare Reinforcement Learning to other data-independent Design Automation approaches to assess potential fields of application for Reinforcement Learning in engineering design. • Method for Reinforcement Learning as an alternative Design Automation approach. • Parametric CAD models as learning environments and requirements as reward functions. • Quantitative Indicators for training effort estimation to enable comparability. • Studies on feasibility, training effort and effects of pre-training. • Identification of potential user groups for Reinforcement Learning. [ABSTRACT FROM AUTHOR]

Current industrial practice does not reflect the opportunities provided by state-of-the-art design automation methods. The limited application of computational methods to support the design process by automating design tasks is caused by the lack of methods for comprehensive design automation task definition. Therefore, potential design automation tasks are not recognized and already deployed solutions lack integration to design practice from a product lifecycle management (PLM) perspective. In response to these shortcomings, this work proposes a method for identification and integration of design automation tasks that features collaborative workshops and enterprise architecture modelling for comprehensive analysis of design processes including its technological environments. The method applies design automation task templates that contextualize the knowledge levels required for design automation task definition and the design process including its technological environments. Evaluation with three industrial cases shows that the method enables efficient identification and integration of potential design automation tasks in a PLM context. The application of knowledge levels in conjunction with enterprise architecture modelling support the identification and validation of the relevant sources of knowledge required for design automation task formalization. Thus, this work contributes by introducing and evaluating a novel method for design automation task definition that brings the opportunities of state-of-the-art design automation methods into line with requirements stemming from design practice and the related PLM. [ABSTRACT FROM AUTHOR]

Abstract: An adaptable methodology for automation application development (AMAAD) is introduced. This development methodology is based on the key concept that design automation (DA) applications are a subset of higher level knowledge-based engineering (KBE) applications, and thus can be developed using a subset of KBE methods. The proposed methodology is largely built upon two existing KBE methodologies: CommonKADS and MOKA, which have become popular models for automating engineering processes. The proposed extension of these methods introduces flexibility to tailor the process for producing automation software to the specific needs of the problem through the specification of a number of attributes. These attributes are linked to subtasks in the key lifecycle phases of application development. This proposed methodology provides a link between KBE and DA applications and provides structure to the application development process. A software tool was written to facilitate the process of identifying the capability needs of an automated solution, and providing detail of the tasks to be followed for its development. [Copyright &y& Elsevier]

Abstract: In the field of aerospace engineering currently a lot of research effort is directed towards the reduction of cruise drag of civil transport aircraft in order to reduce fuel burn, and hence environmental impact and costs. In order to reduce cruise drag, a promising method is under consideration by adjusting, or rather morphing the rear part of the aircraft’s wing during cruise flight. Given the premature state of knowledge of such a design implementation, a knowledge-based computational framework is developed. The purpose of this framework is to allow for an aerodynamic optimization of a section of the wing. The framework is set up in such a way that all relevant design knowledge generated in the process can be captured and used in a subsequent mechanical design process. In this fashion, the complex design process of a novel morphing wing device can be automated to a certain degree. This automation can be used to construct a large number of different feasible and optimized designs with varying boundary conditions of a complex experimental device. This article describes the initial 2-dimensional aerodynamic design step of the morphing device under consideration and how it is implemented in a knowledge-based optimization framework. It describes the initial stage of the development of this tool, as it will be expanded by a number of design steps that each adds more detail to the design in all relevant aspect fields (aerodynamic, structural, actuation, etc.). Ultimately, this tool will be used to obtain a thorough evaluation of a number of different proposed structural solutions and allow for a comparison between them. [Copyright &y& Elsevier]

Abstract: This paper explores novel methodologies for enabling Multidisciplinary Design Optimization (MDO) of complex engineering products. To realize MDO, Knowledge Based Engineering (KBE) is adopted with the aim of achieving design reuse and automation. The aim of the ongoing research at Linköping University is to shift from manual modeling of disposable geometries to Computer Aided Design (CAD) automation by introducing generic high level geometry templates. Instead of repeatedly modeling similar instances of objects, engineers should be able to create more general models that can represent entire classes of objects. The proposed methodology enables utilization of commercial design tools, hence taking industrial feasibility into consideration. The concept of High Level CAD templates (HLCt) will be proposed and discussed as the building blocks of flexible and robust CAD models, which in turn enables high fidelity geometry in the MDO loop. Furthermore, quantification of the terms flexibility and robustness are presented, providing a means to measure the quality of the geometry models. Finally, application examples are presented in which the outlined framework is evaluated. The applications have been chosen from three ongoing research projects aimed at automating the design of transport aircraft, industrial robots, and micro air vehicles. [Copyright &y& Elsevier]

Abstract: Sub-suppliers acting in the supply chain of the car industry have to adopt a strategy of products customization to be at the competitive edge. This is caused by the diversity of client’s requirements that entails customer engineered products with a high level of variety. A business strategy based on engineered-to-order products requires systems for efficient generation of product variants. This also includes a need of decision support in the order preparation process as well as in the quotation preparation process. Decision support that gives access to detailed and accurate information in the quotation preparation enables a high level of product adaptation while ensuring company efficiency. By the introduction of a design automation system, this can be achieved. However, means to manage different application domains, projects, task knowledge and design information are required together with a possibility to perform detailed analyses on the system generated information. The objective of this work is to contribute in that area. Initially, a number of success criteria were identified and explorative work was conducted for the purpose to develop a conceptual model and principles that an application system would reside upon. The result is a framework consisting of an information model and underlying principles to be used when developing a design automation system for quotation preparation. A system founded on the framework supports management and analysis of quotations and product variants. The functionality and usefulness has been demonstrated and validated by a system implementation developed in collaboration with an industry partner. [Copyright &y& Elsevier]

Abstract: Many companies base their business strategy on customized products. To enable a high level of product adaptation in an engineer-to-order approach companies invest time and resources to develop design automation systems. Initially, when implementing a design automation system, the focus is on successfully developing a system that generates design variants based on different customer specifications (i.e. the execution of system embedded knowledge and system output). However, in the long run, two important aspects are the management and maintenance of the knowledge that governs the designs. Further, the increasing emphasis on deploying a holistic view of a product’s properties and functions implies an increasing number of life-cycle requirements. The knowledge to adapt the product to fulfil these requirements should also be used and consequently incorporated into the knowledge-base, allowing for correct decisions to be made. In a system for automated variant design, the implications on the product of these life-cycle requirements have to be expressed as algorithms, production rules and/or computational statements to be intertwined with the design calculations. The number of requirements can be significantly large, and the knowledge scattered over different application systems used for the realisation of the design automation system. This makes it difficult to manage and maintain the system as the product life-cycle environment changes and evolves. In this article, the focus is on the requirements related to manufacturing. For that, an approach for the modelling of manufacturing requirements, supporting both knowledge execution and information management, in systems for automated variant design is introduced. The approach has been applied and refined when developing a design automation system in cooperation with a company to demonstrate and verify the approach’s usability. [Copyright &y& Elsevier]

Abstract: As the field of design automation and generative design systems (GDS) evolve, more emphasis is placed on issues of design evaluation. This paper focus on the presentation of different applications of GENE_ARCH, an evolution-based GDS aimed at helping architects to achieve energy-efficient and sustainable architectural solutions. The system applies goal-oriented design, combining a genetic algorithm (GA) as the search engine, with the DOE2.1E building energy simulation software as the evaluation module. Design evaluation is based on energy spent for heating, cooling, ventilation and artificial lighting in the building, and on sustainability issues like greenhouse gas emissions associated with the embodied energy of construction materials. The GA can work either as a standard GA or as a Pareto GA, for multicriteria search and optimization. In order to provide a broad view of the capabilities of the software, different applications are discussed: (1) standard GA: testing and validating the software; (2) standard GA: incorporation of architecture design intentions, using a building by architect Alvaro Siza; (3) Pareto GA: choice of construction materials, considering cost, building energy use, and embodied energy; (4) Pareto GA: application to Siza’s building, considering thermal and lighting behavior separately; (5) standard GA: shape generation with single objective function; (6) Pareto GA: shape generation with multicriteria evaluation; (7) Pareto GA: application to an urban and housing context. Overall conclusions from the different applications are discussed, as well as current challenges and limitations, and directions for further work. [Copyright &y& Elsevier]

As engineers design, plan, and execute architecture, engineering, and construction (AEC) projects, they construct task-specific views. These views contain information that is structured for their specific tasks. Engineers generally construct these views based on information contained in other engineers'' views. Traditionally, engineers have manually constructed these views using pencils and more recently with computational tools like computer-aided drafting (CAD). Manually constructing and integrating views is often difficult, time-consuming, and error-prone. Newer project model approaches attempt to predetermine and automatically construct many useful task-specific views but these approaches are proving difficult to implement due to the multidisciplinary, constructive, iterative, and unique character of AEC projects. Current project modeling approaches lack adequately simple, formal, generic, expressive methods that engineers can use to automatically construct a new dependent view from information in one or many source views. In this paper, we formalize reusable reasoning modules, called Perspectors, which engineers can use to automatically construct a task-specific engineering view, called a Perspective, from other Perspectives. We motivate and retrospectively validate this approach using engineering test cases from the design and construction of the Walt Disney Concert Hall. Through implementation of a prototype, we give empirical evidence that engineers can select from a potentially small number of pre-defined, reusable, geometric Perspectors, or program new Perspectors, and easily compose them into directed graph structures, called Narratives, to construct and control the integration of useful dependent geometric Perspectives more quickly and accurately than current practice and theory allows. Perspectors are intended to enable engineers from multiple disciplines to engage in novel, automated, and integrated design and analysis by easily yet formally constructing and integrating Perspectives from other Perspectives. [Copyright &y& Elsevier]

A product platform is a design approach for meeting the demand for customizable products. Traditional knowledge-based technologies or systems lack flexibility in supporting both configuration and parameter design of platform-based products. In many cases, customers' requirements and knowledge models both contain incomplete information, and there are complex relations among various solutions, functions and solution parameters in Engineering-To-Order (ETO) products. A knowledge model for the preliminary design of ETO products is presented in this paper, and linkages are established between configuration design knowledge and parameter design procedures. The basis of the knowledge model is the Extended Function-Solutions (EFS) tree, from which design case trees, design modules, constraint checking rules, and module interface templates derive. A corresponding knowledge retrieval and reuse strategy is also presented. It uses an improved fuzzy information axiom to search for the optimal configuration with incomplete information. The parameter design process model of new products then can be generated based on the optimal configuration. The case study demonstrates the knowledge modeling, retrieval and reuse for the preliminary design of open-type crank presses. Moreover, the effectiveness of the methodology is discussed by analyzing the verification approach and the satisfaction of customers' requirements. [ABSTRACT FROM AUTHOR]