NSF Convergence Accelerator Track D: A Community Resource for Innovation in Polymer Technology (CRIPT)

聚合物材料，从服装和个人防护设备到建筑材料和食品包装，是满足我们对食品，住房，健康和交通的基本需求的基础。然而，为下一代产品开发新的聚合物需要数十年的时间，我们必须加快步伐以保持竞争力。为了加速这一过程，该项目正在开发CRIPT，这是一个聚合物数据生态系统，由基于Web的应用程序和云数据库组成，使聚合物科学家能够轻松地查找，存档和与复杂的聚合物数据进行交互。CRIPT内部的人工智能驱动的化学工具和数据驱动的工作流程将使聚合物材料的开发时间减少一个数量级，对每年销售近6000亿美元的聚合物的生产商和买家产生变革性影响。目前，在现有聚合物中搜索是一项艰巨的任务，因为聚合物数据以小而分散的集合存在，使导航成为一个复杂的过程，其中包括统一不同的数据格式和协调元数据，这两方面目前都需要专家干预。CRIPT提供了一个基于新的聚合物特定数据模型的云数据库，该模型同时提供了聚合物科学和工程不同领域的互操作性，同时保留了关键的元数据，使领域专家能够将许多独立记录中的信息关联起来。还提供了一系列受化学启发的人工智能创新，包括基于化学的查询语言，基于图形的模式，保留数据中的时间结构，自动数据验证算法，用于数据摄取的人工智能-人类合作工具，以及将机器集成到数据生态系统中，以增加公平原则，数据信任，该奖项反映了NSF的法定使命，并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。

Polymer materials, ranging from clothing and personal protective equipment to construction materials and food packaging, are fundamental to providing for our basic needs for food, shelter, health, and transportation. However, developing new polymers for next-generation products takes decades, and we must move faster to remain competitive. To accelerate this process, this project is developing CRIPT, a polymer data ecosystem consisting of a web-based application and cloud database that allow polymer scientists to easily find, archive, and interact with complex polymer data. AI-driven chemistry tools and data-driven workflows within CRIPT will reduce the development time for polymer materials by an order of magnitude, creating a transformative impact on both the producers and buyers of the nearly $600 billion of polymers sold each year.Currently, searching among existing polymers is a daunting task because polymer data exists as small, disparate sets, making the navigation a complex process combining the harmonization of different data formats and the reconciliation of metadata, both of which currently require expert intervention. CRIPT offers a cloud database based on a new polymer-specific data model that simultaneously provides interoperability across different domains of polymer science and engineering, while retaining critical metadata that allows domain experts to correlate information across many independent records. A series of chemically-inspired AI innovations, including a chemistry-based query language, a graph-based schema preserving temporal structure in data, algorithms for automatic data validation, AI-human cooperative tools for data ingestion, and the integration of machines into the data ecosystem are also provided to add FAIR principles, trust in data, and ease of use to the system.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.