My team of Design Technologists created prototypes of both the software we designed and also early versions of the hardware. In most cases, when prototyping hardware we created working versions of designs created by the industrial design team, which was not part of our org. But by creating these early working versions, we allowed our UI design team to have a platform for testing their designs on the actual form factors which would eventually ship.

In a few cases, such as the Echo Wall Clock and the Echo Show 10 with motion, my team created the industrial design and the core as well as the functional prototypes and shepherded the products all the way through to ship. In other cases, such as the Echo Spot, we sourced our own component hardware and used it to build the first working versions of the product. We used our prototypes to assess designs, but also to run usability studies and evaluate various technical innovations my team put forward.

On the software side, we created working versions of all of our UI designs. Our interaction design principles called for the device to always be voice-forward, with touchscreen interactions used to supplement rather than dominate over voice. Our goal was to ship a voice device with touch, not a tablet with voice. This was at times a challenging balance to strike, as many team members we should fall back to known design patterns.

We also created numerous tools to allow us to demo Alexa functionality and response to utterances that take a long time to train. We also created tools to make it easier for our leadership team across multiple locations to assess our designs on device.

In a few cases my team innovated new technical solutions which shipped on the product. One example is the person-tracking technology that the Show uses to detect presence, which was based on the work my team did on an internal remote collaboration tool. Another is the Echo Show 10 with Motion, which is based on the robotics studies my team began shortly after the launch of the original Echo Show.

As more and more 1P and 3P teams created Alexa skills we began to see a recurring set of design mistakes from our partners. To address this, my team first created a set of human interface guidelines for Alexa devices. This eventually evolved into a design system we called Pantry.

This team evolved out of our production design and visual QA team, and ended up growing to 4 full time designers.

One of our most pressing challenges in the early days of designing the UI for our device was in understanding how we needed to adjust our design choices to the fundamentally different mode of interaction required for a voice-forward device. Users would not be 2 feet away from the device like they typically are for a phone, tablet or laptop, or 10 feet on a television. After performing some early studies in a prototype kitchen we built using our prototype hardware, we decided to base our UI designs on the requirement that they be readable from 7 feet, in order to allow the device to be voice-forward.

Making this change meant that all of our design choices needed to be adjusted. Instead of points, pixels, or DP, we had to use Arc Minutes as our primary metric when determining the correct size of assets. Arc minutes measure the amount of a user's field of view a visual element uses at a given distance. My user research team had to explain to all designers and stakeholders what Arc Minutes were and how they applied to our designs. We created numerous assets to explain the metric to those who were unfamiliar with it, and we created a tool to allow designers to convert pixels to arc minutes at 2', 7' and 10'. This tool allowed designers to understand how asset sizes needed to adjust to different use cases.