As a Senior Interaction Designer working on Samsung's Gear Icon X earbud headphones, I created prototypes, defined touch based interactions, developed the audio UI, and created specifications for lighting feedback. I also collaborated with industrial designers on ergonomic studies of ears.
https://news.samsung.com/global/interview-how-the-gear-iconx-got-its-fitting-design
In the Fall of 2014, I joined Samsung as an Interaction Designer. One of my earliest accomplishments was hosting a team hackathon that resulted in 16 functional prototypes created within a week for less than $500. Among the prototypes created was a wearable, screenless media player intended to track fitness that I’d designed and built using an Arduino LilyPad and a 3D printer. Our goal was to integrate onboard music with a fitness tracker, with an ambitious stretch goal of incorporating Samsung’s Milk or Spotify.
However, the real breakthrough came from a critical insight during this process. We discovered that when people work out, they mostly focus on listening to music and tracking physical progress. This unique insight illuminated a promising opportunity for Samsung at the intersection of music and fitness, guiding our future product development strategies.
In the Spring of 2015, several months after the hackathon, we received the green light to create a fitness-focused set of headphones for Samsung users. In the years prior, Samsung Design America had made the Level and Level Over headphones, although these required users to carry a phone or laptop to listen to music.
As a team, we saw from the hackathon and supporting market research that there was an opportunity to build a hearable product that tracked steps and enabled users to bring their music along with them.
This was the genesis of the Gear IconX (with integrated flash memory), and these insights further informed the design of the Gear Fit 2 watch, which supported a Spotify app and pairing with the IconX.
The hearing aid market was in a strange place. Hearing aids were highly regulated and government approval added a significant premium to the retail cost putting them out of reach of people who needed them. Many people were and are still suffering from partial hearing loss. If we could provide a similar feature set, and an app that enabled personalized tuning of an audio environment, targeted to people with partial hearing loss, we could create a competitive product.
To better understand this market, I studied audiology, took a hearing test, and attended Project Bar-B-Q where I connected with other audio experts facing similar challenges. Shortly after that, Bose began actively working to support the deregulation of the hearing aid market.
The shape of people’s ear canals are wildly different and our industrial designers needed anthropometric data about inner ears to create an optimal fit for the ear buds. To solve for this, I created inner ear molds of coworkers that we measured to inform the constraints and requirements of our industrial designs. I also did an ergonomic study of earbud adapters.
Since we’d recently worked on a home voice agent for Samsung, voice control was top of mind. We looked into what it would take to bring a standalone voice agent into the Icon X and ultimately determined that it wasn’t worth it. Product tradeoffs involved speed of voice detection, accuracy, breadth of recognized phrases, physical size, and power consumption. For cost and simplicity, we opted instead to enable voice interaction only when tethered to a cell phone.
To validate and showcase design ideas, I white-boarded interaction concepts and used those to build audio and touch-based interaction prototypes in Noodl that showed how a user could navigate the device’s features and accessories. This prototype simulated features such as environmental audio passthrough, music playback, and audio menu navigations when using tap, double-tap, triple-tap, and press-and-hold functionality as input. I recorded voice talent for the onboard responses as well. Wit.ai was our preferred speech recognition platform for this prototype although its usage was deprioritized in favor of robust documentation showing interaction options with recommendations. This is what eventually shipped.