Research and Story Development
The process began with brainstorming different emotions people might feel at a time, the activity they might do when experiencing that emotion, and developing a short narrative/story surrounding that feeling. ​​​​​​​
These written ideas were then turned into visual ideas. Some
ideas were very serious, while others were practically impossible
to pull off. 
Presentation for Feedback
After presenting the concepts in class, I received a lot of great feedback on how to improve the narrative experience, which concepts were strong (with pink stickies beside them) and which were not as strong (with blue stickies beside them).
Revised Ideation
From peer feedback, I revised some of his ideas. While these ideas above were well thought out, I was pushed to further explore the narrative and how emotions can drive the design of an interactive wearable. 
Revised Story Development
I went back to the drawing board to revise some of his stories that could be used within the experience of the wearable. I began to focus on the feeling of being scared. One story developed was about people visiting an abandoned teddy bear factory and having to fight them off with an interactive blaster. Concepts from these stories were then created, as seen below. From feedback, I found that some of these concepts could be felt by the person wearing them but not seen externally, which is a requirement. 
I then took some of these concepts, including new ones from peer feedback, and began conceptualizing them. This included writing other stories and simplifying
the experience.
Revised Concepts and
Block Diagrams
Due to the best feedback coming from concepts 1 and 3 above, I then revised those concepts in ProCreate, where he put the wearables in context, added storyboards, and detailed the design. At the same time, I created block diagrams of electronic components that would make the interaction occur. Peers felt the "Question Hat" was the strongest concept.
Low-Fidelity Prototype 
I began building the low fidelity model with battery packs, Ardunio Redboard Uno, stemma wires, time of flight distance sensor, and ICM sensor. A simple breadboard working model with one light can be seen in the second to the left picture below.  However, the battery back and Ardunio Uno were too big to be on the hat, so a FeatherS2 was needed, along with a lithium chargeable battery. More LEDs were added as well to form the "?" symbol.  The model was presented in class for peer feedback. 
The hat shown was bought online. All rights reserved to the hat itself and respected parties.
Developing a Revised Model
Final revisions from lessons learned in the first prototype were completed. While an attempt to crimp rainbow wireless did not work out, the design and integration of the wires were revised to fit smoothly into the inside of the hat. Also, the LEDs were strategically placed so that the ground wires would all run together as well. Velostat, a black flat material was used to cover the wires and FeatherS2 that were sewn onto the hat. The final model can be seen and presented below at a desk, just like it would be in the context of its use. 
Revised Model

You may also like

Back to Top