CAREER: Toolkits for Digital/Physical Workflows

该项目通过降低在工作流程中使用自动化的门槛，提供工具包来支持设计人员创建包括硬件和软件的系统。虽然自动化可以提高精度和效率，但设置新的自动化工作流程可能成本高昂，需要专业知识和大量的前期工程时间。由于这种成本，自动化在很大程度上用于那些经常使用的任务，这些任务可以收回前期投资，例如在大规模生产的背景下。在较小规模上工作的领域专家或使用利基工作流的专家不能以同样的方式受益，从而阻碍了他们的进展。相比之下，这项研究优先考虑了更广泛的制造从业者群体，他们的不同需求需要广泛的自动化策略。这个项目将帮助实践者在他们的工作中确定自动化的可行机会，贡献开源软件和硬件工具包，实践者可以使用这些工具包来编写他们自己的自动化数字/物理工作流，并组织使用和维护这些技术的社区。这些产出将通过一系列使用这些技术的课程和讲习班以及通过开放源码的发布，为开发以人为中心的创新系统的广泛人员提供帮助，这些系统在其选定的应用领域使用数字/物理工作流程。本研究包括三个主要领域：对跨越数字和物理空间的从业者（如制造商或实验室科学家）的工作进行形成性研究；开发用于自动化物理过程的新设备和硬件模块，包括运动控制和主动末端执行器；以及支持自动化数字/物理工作流创作的最终用户软件，包括支持创建用于自动事件检测的特定任务人工智能（AI）模型的模块。这种以人为中心的方法——用户可以在收集额外数据时改进他们的模型和工作流——利用了从业者在收集过程中对自己数据的理解。通过形成性研究和设计研讨会与不同领域的专家合作，本研究的工具包和系统将推广到多个应用领域。其他人将有机会参与这项技术，并通过自己动手（DIY）生物实验室和创客空间等社区空间的课程和研讨会了解自动化。此外，这项研究将为开源硬件提供设计建议，这些硬件是可复制的，可以由最终用户定制。总之，这些贡献构成了自动化数字/物理工作流程的有价值的工具包和对未来系统设计的影响。该奖项反映了美国国家科学基金会的法定使命，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。

This project contributes toolkits to support designers creating systems that include both hardware and software by lowering the threshold to using automation in their workflows. While automation can increase precision and efficiency, setting up new automated workflows can be costly, requiring expertise and significant up-front engineering time. Because of this cost, automation is largely used for tasks that are sufficiently frequently used that this up-front investment can be recouped, such as in the context of mass production. Domain experts working at smaller scales or with niche workflows cannot benefit in the same ways, hindering their progress. In contrast, this research prioritizes a broader community of fabrication practitioners whose varied requirements demand a broad diversity of automation strategies. This project will help practitioners identify viable opportunities for automation in their work, contribute open-source software and hardware toolkits with which practitioners can author their own automated digital/physical workflows, and organize communities using and maintaining these technologies. These outputs will provide assistance to a wide range of people developing innovative human-centered systems that use digital/physical workflows in their chosen application areas through a series of courses and workshops that will be taught using these technologies and through their open-source release.This research includes three main areas: formative research on the work of practitioners who span digital and physical spaces, such as fabricators or laboratory scientists; the development of novel devices and hardware modules for automating physical processes including motion control and active end effectors; and end-user software that scaffolds the authoring of automated digital/physical workflows, including modules that support the creation of task-specific artificial intelligence (AI) models for automated event detection. This human-centered approach—where users can refine their models and workflows as they collect additional data—leverages practitioners’ understanding of their own data during collection. By collaborating with experts across different domains through formative studies and design workshops, toolkits and systems from this research will generalize to multiple application areas. Other people will have opportunities to engage with this technology and learn about automation through courses and workshops in community spaces like Do-It-Yourself (DIY) biolabs and makerspaces. Furthermore, this research will contribute design recommendations for open-source hardware that is replicable and can be tailored by end-users. Together, these contributions comprise both valuable toolkits for automating digital/physical workflows and implications for future systems design.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.