Research vision · Human–computer interaction
Interactive Tools for Design Reasoning
Interactive design tools have become increasingly capable of helping designers create and realize their ideas. Building on these advances, I investigate how interactive tools can also support designers as they reason through the decisions that shape their designs, from CAD modeling to fabrication.
As design goals, constraints, and possible solutions evolve, designers may encounter overlooked requirements, explore alternatives, or leave decisions open while new knowledge becomes available.
Research question · How can interactive tools support informed design decisions across design and fabrication?
I organize this research agenda around problem and solution reasoning, and the divergent and convergent activities that can occur within each. These directions overlap as designs develop.
Updates & milestones
Presented the TOM assistive technology search engine with Maayan Keren and the TOM Global team at Cornell Tech's AI and Accessibility Summit 2026 ↗.
Siegel PiTech PhD Impact Fellow with TOM Global. Built a multi-source assistive technology search tool. Project details ↗
CAMeleon was accepted to ACM CHI 2026; I attended and presented the work in Barcelona, April 13–17.
OriStitch was featured in Cornell News ↗.
Y-AR and OriStitch were published at the ACM Symposium on Computational Fabrication.
2025 PiTech fellowship work featured in a Cornell spotlight ↗.
PiTech PhD Impact Fellow with NYC Mayor's Office of Climate & Environmental Justice. Flood dashboard ↗
Invited talk at Cornell Tech's XR Collaboratory.
Invited talk at Carnegie Mellon University's Generative Systems for Design class.
Received an XR Collaboratory Prototyping Grant ↗.
Publications, ongoing research & applied projects
Showing all 9 projects
Learners often begin with a familiar fabrication workflow and may lack the knowledge to judge alternatives. Early modeling choices can make it difficult to change direction as their design develops. CAMeleon lets learners compare workflow outcomes for the same model, supporting exploration of trade-offs and the development of informed design judgments.
aCADemy is an AI-based reflective interface that works alongside CAD software. It presents contextual questions about fabrication considerations during modeling, helping designers reflect on decisions while retaining control over how their designs develop.
FuzzyCAD lets collaborators express unresolved design decisions directly on CAD geometry, contribute constraints and identify competing requirements. The project examines how collaborators with different expertise can progressively develop these decisions across asynchronous handoffs.
People seeking assistive technology often describe everyday challenges without knowing the technical names of relevant solutions. Developed with TOM Global, this tool connects conversational intake to searches across assistive technology catalogs, open-source designs, and commercial products. It helps staff assess whether a result can be used as-is, adapted, or turned into a new maker challenge when existing solutions do not fit.
Developed with the NYC Mayor's Office of Climate & Environmental Justice, this interactive dashboard presents flood sensor readings, historical events, and localized monitoring information in an accessible format. It helps New York City residents explore flood-related data and better understand conditions relevant to their neighborhoods.
Y-AR is a mixed-reality CAD tool for creating three-dimensional wire structures that connect to objects in the physical environment. Its interface incorporates springs as design primitives, both to apply forces that hold objects and to accommodate dimensional inaccuracies from mid-air modeling and measurement.
OriStitch transforms existing flat fabrics into self-folding three-dimensional structures through computational embroidery. Functional threads are stitched in specific patterns and then activated with heat to achieve a target form. The workflow works with a range of materials, including leather, woven fabric, and denim.
Multi-Self is a BCI–VR design tool that provides real-time visual feedback about designers' emotional responses to support metacognitive monitoring. A pilot study with 24 participants reported increased design exploration and self-awareness, while responses to the accuracy of the emotional feedback were mixed.
Emotive Room explores how an interactive architectural environment can respond to people's emotional states through real-time brain–computer interaction. By changing spatial color and form in response to stress-related feedback, the project investigates responsive environments that may support calmness and well-being.
Shuo Feng · Ph.D. Student, Information Science · Cornell Tech