SNM: Manufacturing Autonomy for Directed Evolution of Materials (MADE-Materials) for Robust, Scalable Nanomanufacturing

尖端材料的开发和制造通常涉及耗时的材料和工艺设计阶段，随后是大量的样品测试，以调整工艺条件，因为制造规模从实验室扩大到试点计划，再到工业工艺规模。这些不同的步骤提高了将新的和改进的材料引入工业管道的成本障碍，并增加了国内制造先进纳米材料的成本。幸运的是，最近在数值模拟、增材制造和材料快速测试方面的发展表明，材料开发和纳米制造的新方法，以前不同且耗时的阶段可以几乎瞬间进行，以达到最佳的材料结构及其制造的工艺条件。该奖项的重点是通过以下方式改造传统的开环纳米结构材料合成：1)使用通用的3d打印方法来制造这些纳米材料和纳米结构；2)将材料特性表征直接纳入打印过程；3)使用人工智能（AI）算法来调整飞行过程条件以实现所需的材料特性。这些概念和组成部分将被整合到技术课程、实践研究机会和面向广大学生和公众的推广研讨会中。合作伙伴计划通过与当地博物馆的合作，以及课程和课外活动，利用现有的推广和教育活动。该方法和框架是对过程建模、材料合成和表征以及系统设计的研究，以自主发现新的材料结构，减少制造缺陷和不确定性。这个人工智能框架将“理解”工艺-结构-性能关系、制造约束，以及更重要的是，材料性能和制造质量的统计变化。这项研究的智力价值在于发现了通用的纳米制造工具和原料，具有监督遗传算法，可以通过寻找替代设计来自主纠正缺陷并补偿先天的制造不准确性；这与标准工具形成对比，标准工具将不确定性（例如环境控制）降至最低，或依赖于人工干预的制造后表征。该框架将使用纳米级增材制造（AM）作为基本制造工具和纳米结构超材料作为应用进行测试。通过这种方法制备的范式和纳米级超材料对集成系统的可扩展纳米制造具有深远的影响。自主进化参数以满足构造规范的系统范例可扩展到宏观增材制造和制药领域，其中工艺参数空间和可用化学物质很大，设计并不直观。增材纳米制造有可能改变超材料设计，实现多种材料的三维设计，创造复杂的复合元结构。

The development and manufacturing of cutting edge materials typically involves time-consuming materials and process design phases, followed by extensive testing of samples to adjust process conditions as the manufacturing scales up from the lab, to pilot plan, to industrial process scale. These distinct steps drive up the cost barrier to introduction of new and improved materials into the industrial pipeline and increase the cost of domestically manufactured advanced nanomaterials. Fortunately, recent developments in numerical modeling, additive manufacturing, and rapid testing of materials suggest that a new approach to material development and nanomanufacturing, where the previously distinct and time-consuming phases could be carried out nearly instantaneously to arrive at optimal material structure as well as process conditions for its manufacture. The focus of this award is to revamp the traditional, open-loop synthesis of nanostructured materials by: 1) using a versatile 3-D printing approach to manufacture these nanomaterials and nanostructures, 2) incorporate material property characterization directly into the printing process, and 3) use an artificial intelligence (AI) algorithm to adjust on the fly process conditions to achieve desired material properties. These concepts and components will be integrated into technical coursework, hands-on research opportunities, and outreach workshops to a broad range of students and the public. The co-PIs plan to leverage existing outreach and educational activities through their group's collaboration with a local museum, as well as curricular and extracurricular activities. The approach and framework is an investigation of process modeling, materials synthesis and characterization, and system design to autonomously discover new material configurations and reduce manufacturing defects and uncertainty. This AI framework will "understand" process-structure-property relationships, manufacturing constraints, and, importantly, statistical variations in material properties and manufacturing quality. The intellectual merit of this study is the discovery of general nanomanufacturing tools and feedstocks, with supervisory genetic algorithms, that autonomously correct for defects and compensate for innate manufacturing inaccuracies by a search for alternative designs; this is in contrast to standard tools that minimize uncertainty (e.g. environmental controls) or rely on post-fabrication characterization with human intervention. The framework will be tested using nanoscale additive manufacturing (AM) as the fundamental manufacturing tool and nanostructured metamaterials as the application. The paradigm and nanoscale metamaterials made via this approach have far-reaching impacts on scalable nanomanufacturing for integrated systems. The paradigm of systems that autonomously evolve parameters to meet construct specifications is extensible to macroscale additive manufacturing and pharmaceuticals where the process parameter space and chemistries available is vast, and design is not intuitive. Additive nanomanufacturing has the potential to transform metamaterial design by enabling design in 3-dimensions (3D) with multiple materials, creating complex composite metastructures.