CAREER: Toolkits for Digital/Physical Workflows

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.