Planning Grant: Engineering Research Center for Resilient AI Network (RAIN) for next-generation manufacturing

规划资助：下一代制造弹性人工智能网络（RAIN）工程研究中心

• 批准号: 2124295
• 负责人: Ali Shakouri
• 金额: $ 10万
• 依托单位: Purdue University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-08-01 至 2022-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2124295&HistoricalAwards=false
• 关键词: Planning; Grant; Engineering; Research; Center

There is an increasing digital divide making it difficult for small and medium manufacturers to be globally competitive. Although they are the lifeblood of well-paying jobs in rural communities, many small and medium manufacturers throughout the U.S. do not have access to the large datasets needed to train classical artificial intelligence / machine learning (AI/ML) algorithms. Progress in the U.S. is currently hampered by this digital-divide as well as a fragmented ecosystem that obviates rapid scale-up. Unfortunately, the rigidity of this system provides little safety net, as shown during the COVID-19 pandemic and the massive electricity blackout in Texas. This planning grant will help develop a nationwide engineering center that is trusted and empowered by industry partners of all sizes. The vision of the proposed Resilient AI Network (RAIN) for next-generation manufacturing ERC is to empower U.S. manufacturers to increase resiliency and international competitiveness by successfully adopting privacy-preserving, scalable, certify-as-you-build AI/ML practices to erase the digital-divide and accrue the benefit of economy-of-scale without centralization. This new manufacturing framework/paradigm based on the fundamental reimagination of AI/ML techniques will empower U.S. manufacturers of all sizes to lead among global competitors. The planning grant will help foster the partnership with community leaders, K-12 and colleges to develop a plan to attract a new diverse engineering workforce gaining insights from AI and data analytic tools and equipped to bring new products to the market.RAIN will enable distributed edge-AI tools integrating inline imaging, sound, vibration, and multi-sensor analytics. This can significantly improve manufacturing quality control, and machine-up time in many industries. AI-commons optimizing privacy-utility tradeoffs for industry cooperatives will be studied. Human-centered design is essential for resilient and agile manufacturing as new products or changes in supply-chain introduce new conditions requiring human input. The RAIN team brings together experts from AI and computer systems, human-centered design, policy, privacy, and engineering education as well as those with extensive experience in manufacturing scale up, testbeds, and life-cycle environmental and economic analysis. A key focus of this planning grant will be bolstering RAIN’s partnerships, resources, and support among our broad range of stakeholders. A central goal will be enhancing RAIN’s industry partnerships, which require a trusting relationship since our vision is high risk but with a high reward. With this planning grant, RAIN will expand our number of trusted industry partnerships through three key methods: (1) sharing capacity-building and value creation responsibilities among all our university partners by expanding existing industry-university relationships through networking that brings together stakeholders; (2) cultivating new relationships with small and medium manufacturers; and (3) hosting workshops with appropriate stakeholders to gain buy-in from manufacturers to issues of intellectual properties (IP), secure a common platform, and promote workforce development opportunities. RAIN will broaden its activities by developing plans for recruiting and accessing an educated, diverse workforce, as well as promoting training opportunities for current employees within our manufacturing demonstration facilities. RAIN will be implementing a comprehensive plan for promoting a diverse and inclusive team.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.

数字鸿沟日益扩大，使中小型制造商难以在全球范围内竞争。尽管它们是农村社区高薪工作的命脉，但美国各地的许多中小型制造商无法访问训练经典人工智能/机器学习（AI/ML）算法所需的大型数据集。目前，美国的进步受到数字鸿沟以及碎片化生态系统的阻碍，无法快速扩大规模。不幸的是，这一制度的僵化几乎没有提供安全网，正如2019冠状病毒病大流行和德克萨斯州大规模停电期间所显示的那样。这项规划拨款将帮助开发一个全国性的工程中心，该中心受到各种规模的行业合作伙伴的信任和授权。针对下一代制造业ERC提出的弹性人工智能网络（RAIN）的愿景是使美国制造商能够通过成功采用隐私保护，可扩展，证书即建的AI/ML实践来提高弹性和国际竞争力，以消除数字鸿沟并在没有集中化的情况下获得规模经济的好处。这种基于人工智能/机器学习技术的新制造框架/范式将使各种规模的美国制造商在全球竞争对手中处于领先地位。规划拨款将有助于促进与社区领袖、K-12和大学的合作，以制定一项计划，吸引新的多元化工程师队伍，从人工智能和数据分析工具中获得见解，并有能力将新产品推向市场。RAIN将实现集成在线成像、声音、振动和多传感器分析的分布式边缘人工智能工具。这可以显著改善许多行业的制造质量控制和机器启动时间。将研究人工智能共享优化行业合作社的隐私-效用权衡。以人为本的设计对于弹性和敏捷制造至关重要，因为新产品或供应链的变化会引入需要人工输入的新条件。RAIN团队汇集了来自人工智能和计算机系统、以人为本的设计、政策、隐私和工程教育的专家，以及在制造规模扩大、试验台和生命周期环境和经济分析方面拥有丰富经验的专家。这项规划赠款的一个重点将是加强RAIN的伙伴关系，资源和我们广泛的利益相关者之间的支持。一个中心目标将是加强RAIN的行业合作伙伴关系，这需要一个信任的关系，因为我们的愿景是高风险，但高回报。有了这笔规划补助金，RAIN将通过三个关键方法扩大我们值得信赖的行业合作伙伴的数量：（1）通过将利益相关者聚集在一起的网络，扩大现有的产学关系，在我们所有的大学合作伙伴之间分担能力建设和价值创造的责任;（2）与中小制造商培养新的关系;（3）与其他大学合作伙伴建立合作关系。以及（3）与适当的利益相关者举办研讨会，以获得制造商对知识产权问题的支持，确保共同平台，并促进劳动力发展机会。RAIN将通过制定招聘和获得受过教育的多元化劳动力的计划来扩大其活动，并在我们的制造示范设施内为现有员工提供培训机会。RAIN将实施一项全面的计划，以促进一个多元化和包容性的团队。该奖项反映了NSF的法定使命，并已被认为是值得通过使用基金会的智力价值和更广泛的影响审查标准进行评估的支持。