FMNet: Building a Network to create the Workforce Foundation, Actionable Roadmap, and Infrastructure Design to Integrate Data Science, AI, and Predictive Analytics throughout Bioma

项目摘要-FMNet：建立网络以创建劳动力基础、可操作路线图和基础设施设计，以在整个生物制造过程中整合数据科学、人工智能和预测分析生物制造，特别是细胞和基因治疗制造的变革性领域，在FDA首次批准和新兴生物制药药物的巨大前景之后，已迅速成为美国和世界各地生物技术和制药行业最关键的部门之一。一个由行业、政府、临床和学术领袖组成的公私联盟认为，缺乏熟练的细胞制造劳动力，以及缺乏数据分析和数据科学方面的培训，是加速将有前途的疗法转化为产品的两大障碍。数据分析、人工智能(AI)和预测建模可以显著加快进展并解决目前在细胞和基因治疗制造方面遇到的技术缺陷；然而，生物制造专家、数据科学家和AI专家之间存在巨大的知识差距。显然，这个行业的成功需要涵盖细胞培养和细胞生物学、生物过程工程、传感器和自动化、制造和供应链物流、监管知识以及最重要的--数据科学、人工智能和预测分析--的专业知识的融合。然而，目前还不存在这样的多学科培训平台。这就是为什么迫切需要一个未来制造网络来在这些领域之间建立一座桥梁，并为知识共享、传播以及最终课程认证要求创造必要的基础。将细胞作为一种治疗产品制造带来了复杂的挑战，与制药和生物技术行业目前所经历的挑战截然不同。首先，产品(细胞)是一个“活的、会呼吸的”实体，它的性质和功能可以随着每次操作而改变，这需要一个全新的大规模制造和质量控制范例。其次，整个领域几乎没有标准化、自动化和在线质量控制，用于细胞收集、细胞特性和处理、细胞身份标记、效力分析和存储解决方案。如果没有能力实施基于关键质量属性、关键工艺参数和按设计总体质量的制造原则的制造工艺，就不可能实现大规模、可重复、低成本地生产高质量和安全的细胞，细胞疗法的影响将仍然很小。拟议的人工智能：生物制造网络的结果将深刻影响生物制造领域未来的工作，重要的是，有助于重新培训当前的劳动力，重点放在人工智能、数据分析和预测建模方面。特别是，该网络专注于确定教育水平(从高中到研究生水平)的关键技能，这将确保满足当前和未来的大量劳动力需求，并填补美国境内熟练、高级和未来的制造业工作岗位。通过该网络开发的能力模型、教学工具、教育研究和培训材料将广泛适用于生物制药制造行业--除了细胞和基因治疗行业。数据科学和人工智能方面也将适用于整个生物技术和制药行业，从而为未来的求职者和再培训的劳动力创造一个广阔的机会空间，同时也从教育的最早阶段招聘和准备一批更年轻、多样化的受训人员。生成的路线图将作为世界各地团体的开源指导文件，产生类似于NCMC电池制造路线图(从现在到2030年)的影响。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

PROJECT ABSTRACT- FMNet: Building a Network to create the Workforce Foundation, Actionable Roadmap, and Infrastructure Design to Integrate Data Science, AI, and Predictive Analytics throughout BiomanufacturingBiomanufacturing, especially the transformative areas of cell and gene-therapy manufacturing, has rapidly become one of the most critical sectors of the biotech and pharma industry in the United States and around the world after the first FDA approvals and the tremendous promise of emerging biopharmaceutical drugs. A public-private consortium of industry, government, clinical, and academic leaders identified lack of skilled cell-manufacturing workforce and lack of training in data analytics and data sciences, as two major barriers for the acceleration of promising therapies into products. Data analytics, Artificial Intelligence (AI), and predictive modeling can significantly accelerate progress and address the technical shortfalls currently experienced in cell and gene-therapy manufacturing; however, there is vast knowledge gap between biomanufacturing experts, data scientists, and AI experts. It is clear that a convergence of expertise spanning cell-culture and cell-biology, bioprocess engineering, sensors and automation, manufacturing and supply chain logistics, regulatory knowledge, and most importantly - data sciences, AI, and predictive analytics, is needed for the success of this industry. However, such a multi-disciplinary training platform does not currently exist. This is why a Future Manufacturing Network is urgently needed to create a bridge between these fields and create the building blocks necessary for knowledge sharing, dissemination, and ultimately, course certification requirements. Manufacturing cells as a therapeutic product poses complex challenges, quite different from those currently experienced by the pharmaceutical and biotech industry. First, the product (cells) is a “living, breathing” entity whose properties and function can change with every manipulation, requiring a whole new paradigm for large-scale manufacturing and quality control. Second, very little standardization, automation, and in-line quality control exists across the field for cell collection, cell characterization and processing, cell identity markers, potency assays, and storage solutions. Without the ability to implement manufacturing processes based on manufacturing principles of Critical Quality Attributes, Critical Process Paramaters, and overall Quality by Design, large-scale, reproducible, low-cost production of high-quality and safe cells cannot be achieved, and the impact of cell therapies will remain narrow. The outcomes of the proposed AI:Biomanufacturing Network will deeply impact both future jobs in biomanufacturing and, importantly, help retrain the current workforce with a specific focus in AI, data analytics and predictive modeling. Particularly, the Network is focused on identifying the key skillsets across education levels (high school through graduate levels), which will ensure meeting the current and future large workforce demands and fill skilled, advanced and future manufacturing jobs within the US. The competency model, teaching tools, educational research, and training materials developed through this Network will be broadly applicable to the biologics manufacturing industry – in addition to the cell and gene therapy industry. The data science and AI aspects will also be applicable to the biotechnology and pharmaceutical industry in general – thereby creating a broad opportunity space for the future job seekers and the retrained workforce, while also recruiting and preparing a younger, diverse group of trainees from the earliest stages of their education. The roadmap generated will serve as an open source guidance document for groups around the world, creating an impact similar to the NCMC cell manufacturing roadmap (now through 2030).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.