Convergence Accelerator Phase I (RAISE): Product Design and Manufacturing Graph-As-A-Service

NSF融合加速器支持以团队为基础的多学科努力，以应对国家重要性的挑战，并在不久的将来展示可交付成果的潜力。这个融合加速器第一阶段项目的更广泛的影响和潜在的社会效益是通过降低开发人员和设计师参与产品设计和制造活动的障碍来促进创新。该项目将启动一个工具的开发，在该工具中，算法设计机器人将协助和协作帮助新手用户设计个性化的消费产品。第一阶段融合加速器项目的目标是汇集工程、认知科学和计算机科学领域的专家团队，沿着包括行业、非营利组织和政府科学家在内的更广泛的利益相关者，为产品设计和制造中的开放知识网络图（称为M-OKN）的设计和实施构建技术路线图。未来用户通过客户发现过程提供的投入将为项目活动提供信息。该项目的长期目标是降低学生，企业家甚至专家设计新产品的障碍，而不必依赖于制造业的丰富经验和主题专业知识。我们的最终目标是开发一种工具，帮助提高美国在先进制造业中的竞争力。我们周围的每一件由人类生产的物品都需要大量的设计工作，无论是日常消费品还是大型复杂的工程系统。然而，设计产品和制造产品所需的大部分知识主要存在于人类的头脑中-经过多年的经验发展或以书籍，文章，博客和在线视频的形式发布。这些知识通常只能通过在产品工程领域的深入研究获得，或者需要广泛的研究和试错。该项目的目标是通过开发一个产品设计和制造知识网络，将三维产品模型信息（包括部件之间的视觉关系）与基础元数据（包括详细的部件信息和与部件制造相关的数据）联系起来，使设计和产品开发更容易获得。大量非结构化、结构化和半分类数据将从网络上收集，并将使用新的几何深度学习和自然语言处理算法进行进一步处理和分析。收集的数据和计算工具将被完善，以创建一个知识网络，为设计机器人提供动力，这些机器人可以回答查询并协助设计。最终，研究界、学生和公众将获得这个开源知识网络的访问权，这有助于加速制造业创新和创业。此外，M-OKN项目还将推动设计和制造科学与信息学和人工智能之间的新融合研究。该奖项反映了NSF的法定使命，通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The NSF Convergence Accelerator supports team-based, multidisciplinary efforts that address challenges of national importance and show potential for deliverables in the near future. The broader impact and potential societal benefit of this Convergence Accelerator Phase I project is to facilitate innovation by lowering the barriers for developers and designers to engage in product design and manufacturing activities. This project will initiate the development of a tool in which algorithmic design bots will assist and collaboratively aid a novice user to design personalized consumer products. The goal of this Phase-I Convergence Accelerator project is to bring together a team of experts ranging from engineering, cognitive sciences, and computer sciences, along with broader stakeholders including industry, non-profits, and government scientists to build a technology roadmap for the design and implementation of an Open Knowledge Network Graph in Product Design and Manufacturing (called M-OKN). Project activities will be informed by input provided by future users through a customer discovery process. The long-term goal of this project is to lower the barriers for students, entrepreneurs, and even experts to design new products without having to rely on significant experience and subject matter expertise in manufacturing. The end goal is to develop a tool that can help improve U.S. competitiveness in advanced manufacturing.Every object around us that has been produced by humans required extensive design effort, whether it is an everyday consumer product or a large and complex engineered system. Yet, much of the knowledge required to design products and to manufacture them resides primarily in human minds - developed over years of experience or published in the form of books, articles, blogs, and online videos. Such knowledge can often only be gained by intensive study in fields of product engineering or requires extensive research and trial and error. The goal of this project is to make design and product development more accessible by developing a product design and manufacturing knowledge network that links 3-D product model information, including the visual relationships between parts, with underlying meta-data that includes detailed part information and data associated with the manufacturing of the part. A large amount of unstructured, structured, and semi-categorized data will be collected from the web and will be further processed and analyzed using novel geometric deep learning and natural language processing algorithms. The collected data and computational tools will be refined to create a knowledge network that powers design bots that can answer queries and assist design. Ultimately the research community, students, and the general public will obtain access to this open-sourced knowledge network which can help accelerate manufacturing innovation and entrepreneurship. In addition, the M-OKN project will advance new convergence research at the interface of design and manufacturing sciences with informatics and artificial intelligence.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.