Assuring the quality of design descriptions through the use of design configuration spaces

今天的全球供应网络的成功取决于设计说明(包括设计意图和形状定义)的高效和有效的沟通，这些说明符合向市场交付产品所涉及的各种工程职能、流程和供应商的要求和能力。技术产品数据包用于提供这些设计说明。在最近的一次行业峰会上，波音公司的一位代表指出，创造产品所需的大约40%的技术数据位于技术产品数据包中形状定义之外。该项目的重点是材料清单(BOM)，它们是形状定义和40%的非形状相关产品数据的组成部分。BOM是基本的，因为它们充当集成器：调整详细的设计描述，以适应特定工程流程的需求。在保持技术数据包的内部一致性的同时重新配置BOM的能力(其中所有BOM配置都是完整的并且彼此兼容)是一个主要挑战。这项建议建立在一项可行性研究的基础上，该研究探索了使用嵌入*将多个BOM与单个设计描述相关联。从工程设计的角度来看，基于与四家当地中小企业的讨论以及与劳斯莱斯燃烧系统相关的案例研究，我们发现了一个迫切的行业需求，即能够将多个BOM与一个或多个设计描述相关联。这一需求一直被隐藏，因为当前的设计技术倾向于将BOM包含在专有数据表示中。然而，工程师使用BOM和其他设计结构来调整设计描述以满足特定目的。为此，需要新的设计技术，使BOM和其他设计结构可供工程师直接使用。从设计技术的角度，我们已经证明了超立方格可以作为计算空间，在其中可以重新配置BOM。然而，生成的晶格是巨大的，尽管在与利兹高级研究计算团队协商后，我们在晶格生成速度方面取得了数百倍以上的改进，但这个问题本质上仍然是指数级的。对于这个项目，格子将保留在背景中，作为技术设备的一部分。从组织心理学的角度来看，重新配置BoM的能力为在产品开发系统中管理工程知识的新方法创造了机会，这些方法考虑了人类和组织行为以及个人偏好。该项目的目标是建立理论基础，并通过一系列可共享的软件原型进行验证，以实现BOM的重新配置。软件原型将被设计为供学术和工业用户使用，以使用他们自己的数据进行实验，并建立对此类设计工具所需的各种功能的理解。这将使公司能够更好地为其IT系统供应商详细说明其长期信息技术需求。将使用分阶段的软件工程过程，并大约每隔六个月发布一系列开放源码原型。这将为研究团队内部有意义的互动创造机会，并让行业合作伙伴及早获得研究机会，并有机会影响研究方向。同时，通过与行业合作伙伴和其他学科的同事合作开发案例研究，我们将加深对工程设计过程中出现的其他类型设计结构的了解，并发展跨学科的学习机会。*嵌入是一种数学机制，允许构造的一个实例叠加到另一个实例上。

The success of today's global supply networks depends on the efficient and effective communication of design descriptions (including design intent and shape definitions) that suit the requirements and capabilities of the wide range of engineering functions, processes and suppliers involved in the delivery of products to markets. Technical product data packages are used to provide these design descriptions. At a recent industry summit, a representative of Boeing noted that some 40% of the technical data needed to create a product resides outside the shape definitions in the technical product data package. The focus of this project is on the Bills of Materials (BoMs) that are integral parts of both shape definitions and the 40% of non-shape related product data. BoMs are fundamental because they act as integrators: adapting detailed design descriptions to suit the needs of particular engineering processes. The ability to reconfigure BoMs while maintaining internal consistency of the technical data package (where all BoM configurations are complete and compatible with each other) is a major challenge. This proposal builds on a feasibility study that explored the use of embedding* to associate multiple BoMs with a single design description. From an engineering design perspective, based on discussions with four local SMEs and work on a case study related to a Rolls-Royce combustion system, we uncovered an urgent industry need to be able to associate multiple BoMs with one or more design descriptions. This need has remained hidden because current design technologies tend to subsume BoMs in proprietary data representations. However, engineers use BoMs and other design structures to adapt design descriptions for specific purposes. For this reason, new design technologies are needed that make BoMs and other design structures available for engineers to work with directly. From a design technology perspective, we have demonstrated that hypercube lattices can act as computational spaces within which BoMs can be reconfigured. However, the generated lattices are vast and, although we made in excess of hundred-fold improvements in the speed of lattice generation after consultation with the Leeds Advanced Research Computing team, the problem remains exponential in nature. For this project, the lattices will remain in the background, as a part of the technical apparatus. From an organisational psychology perspective, the ability to reconfigure BoMs creates opportunities for new ways of managing engineering knowledge in product development systems that take account of human and organisational behaviours, and individual preferences. The goal of this project is to establish theoretical foundations, validated through a series of sharable software prototypes, to enable the reconfiguration of BoMs. The software prototypes will be designed for use by academic and industrial users to experiment with their own data and build understanding of the kinds of functionality required in such design tools. This will allow companies to better specify their long term information technology requirements for their IT system providers. A staged software engineering process will be used and a series of open source prototypes published at roughly six month intervals. This will create opportunities for meaningful interactions within the research team, and give industry partners early access to the research and opportunities to influence the research direction. In parallel, through the development of case studies in collaboration with industry partners and colleagues in other disciplines, we will build understanding of other types of design structure that occur in engineering design processes and develop cross-disciplinary learning opportunities. * Embedding is a mathematical mechanism that allows one instance of a construct to be superimposed on another.