Projects Index

[The Brief]

Most smart wearables collect physiological data but rely on screens and metrics that increase cognitive load. We explored how smart wearables should move beyond passive tracking to become responsive companions that sense, interpret, and communicate the body's needs in intuitive ways.

[The Methodology]

We designed vCare as a closed-loop wearable that continuously senses physiological signals and translates them into subtle heat, vibration, and light feedback through smart textiles. Instead of displaying raw data, the system delivers embodied cues while providing optional real-time analytics through a companion interface.

[The Outcome]

The resulting prototype demonstrates how smart wearables can integrate continuous sensing, contextual interpretation, and adaptive feedback to support stress awareness and self-regulation, shifting from data collection to meaningful interaction.

[The Brief]

Sit-to-stand is one of the most mechanically demanding daily movements, requiring significant lower-limb strength and balance. For older adults and individuals with reduced mobility, this transition often becomes a major source of fatigue and instability. We set out to understand the biomechanics of sit-to-stand movement and translate quantitative human-motion data into a wearable assistive system that provides support only when it is needed.

[The Methodology]

We developed a research-driven workflow combining biomechanical analysis, human-subject testing, and wearable prototyping. A custom experimental rig was built to evaluate multiple bar positions while recording force distribution, EMG activity, and body kinematics throughout the sit-to-stand cycle. These findings informed the design of a lightweight passive exoskeleton that aligns with natural joint rotation and stores mechanical energy during flexion before assisting users through the highest-demand transition phase. The final system integrates ergonomic geometry, compliant materials, and passive mechanical assistance into a wearable structure.

[The Outcome]

The resulting wearable demonstrates how biomechanical research can directly inform assistive product design. By identifying peak mechanical demand during the momentum-transfer phase, the passive exoskeleton provides targeted support without motors or external power. The project establishes a human-centered framework that bridges motion analysis, wearable engineering, and ergonomic design, offering a scalable approach for improving mobility, reducing muscular effort, and supporting independent movement.

[The Brief]

I wanted to explore how advanced biomechanical research could be translated into an intuitive digital experience. Rather than relying on technical specifications alone, my goal was to present complex performance data through immersive visuals that make the science behind adaptive footwear engaging, accessible, and easy to understand.

[The Methodology]

I developed a "Science Through Visualization" approach, combining high-fidelity 3D rendering with interactive biomechanical simulations. I designed a layered information system that allows users to explore gait cycles, muscle activation, morphology, and performance metrics progressively, making technical content feel clear without sacrificing depth.

[The Outcome]

I transformed complex biomechanical concepts into a compelling visual narrative that balances scientific accuracy with premium design. The final experience demonstrates how thoughtful interaction design can simplify sophisticated research while creating a memorable product showcase for athletes, designers, and stakeholders.

[The Brief]

We explored how AI could make the invisible experience of smell more understandable through interaction design. Inspired by the traditional Japanese Genji-kō incense game, we set out to create a collaborative human–AI experience that combines biosensing, machine learning, and conversation. Our goal was to transform the abstract nature of olfactory perception into an engaging, reflective, and culturally grounded interaction.

[The Methodology]

We developed an "AI as a Co-Smelling Partner" framework that integrates olfactory sensing hardware, an LLM-powered conversational interface, and a mobile application. Rather than identifying scents for users, the system encourages comparison, dialogue, and reflection through multiple rounds of guided interaction. Inspired by Genji-mon patterns, we visualized participants' scent judgments to reveal both the similarities and differences between human and AI perception.

[The Outcome]

The resulting experience demonstrates how AI can extend human perception beyond traditional visual and language-based interactions. By combining biosensing, conversational AI, and interaction design, we introduced a new approach to collaborative olfactory experiences that encourages curiosity, discussion, and self-reflection. The project was presented at CHI 2026, highlighting the potential of AI as a partner for exploring human sensory perception rather than simply automating it.

[The Brief]

Traditional robotic systems are often rigid, mechanically complex, and difficult to adapt to dynamic environments. We explored how pneumatic soft robotics could provide a lightweight, compliant alternative by combining flexible materials with modular fabrication. The challenge was to design a soft robotic structure that could be rapidly fabricated, assembled, and actuated while maintaining structural stability and repeatable motion.

[The Methodology]

We developed a modular pneumatic soft robotic system that integrates computational design, digital fabrication, and physical assembly into a unified workflow. Each inflatable component was optimized for controlled deformation and fabricated using precision-cut flexible materials before being assembled into a scalable structural network. By combining air-driven actuation with lightweight modular joints, we transformed flat fabricated components into three-dimensional adaptive structures capable of programmable movement and reconfiguration.

[The Outcome]

The resulting prototype demonstrates a scalable approach to pneumatic soft robotics that emphasizes rapid fabrication, modular assembly, and adaptable motion. The lightweight structure can be assembled efficiently while maintaining consistent pneumatic performance across multiple configurations. Together, the project establishes a flexible design framework for future applications in adaptive architecture, wearable technologies, and next-generation soft robotic systems.

[The Brief]

Adaptive Structural Reimagining translates personal search histories into responsive physical structures that visualize digital identity. Python and LDA analysis identify recurring interests, while Arduino-connected sensors convert the resulting data into changing balloons and deformable poly-line structures. Kangaroo simulations, skeletal analysis, and body-language studies inform an adaptive tensile system that reveals the fluid, intimate, and multifaceted nature of online identities.

[The Brief]

The research question asks: How can we visually synthesize ambiguous EEG signals with environmental anchors to differentiate positive and negative high-arousal states, specifically excitement versus fear? Furthermore, how does the sudden integration of audio onto a visual baseline catalyze this change? This is built on the foundation that physiological signals alone are semantically thin and lack a simple one-to-one mapping with our psychological states.

[The Brief]

Symmetry Grammar examines the geometric rules behind the Alhambra’s intricate patterns. Created during the Digital Futures Shape Grammar Workshop, the project uses personalized GH tool to visual computation and the Shape Machine to analyze symmetry, develop transformation rules, and generate new pattern variations. It bridges historic ornamentation and contemporary computational design.

[The Brief]

Evolve from Singularity explores life’s evolution within a collapsing universe—from simple organisms to intelligent alien civilizations. HTC Vive motion tracking captures natural movement, Grasshopper transforms the data into evolving parametric forms, and Blender brings the resulting creatures to life through animation. The project combines computation, biological behavior, and speculative design to imagine how cosmic rebirth could generate new forms of life.