Thread by @bartonsmith, For the millions of Americans driving older cars, I'm excited to announce [...]

For the millions of Americans driving older cars, I'm excited to announce our smart player called Car Thing.

I'm so grateful to have led this project over the last while and am thankful for my co-designers, Karl, Yoon and Richard, and everyone else who contributed!

I've been pretty nervous about this launch, in part from Covid life, in part cos I focus too much on mistakes, and in part from the backlog of JIRA tickets we didn't get to.

But my therapist says I need to practice focusing on positives, so here are some things I'm proud of...

Building a product is hard. Building a HW/SW product is hard. Building a company's first hardware product is hard. Building a product during a pandemic is hard. Building a product with a team in a different timezone is hard.

It's pretty dope that we made this happen!

Working with dedicated hardware brings its own unique challenges. You can't rely on all of the intuition you've gained from mobile. Building the whole app and its logic in @FacebookOrigami helped us make rapid decisions.

I love this tool and the feeling of new skills emerging.

Conveniently, Car Thing is the same size as an iPhone X's screen. This made it much easier to use and share prototypes, and to understand the scale and distance we were designing for.

I was also happy with our @figmadesign specs, process, and communication with engineers. Being OCD was especially helpful when you only have a couple of hours of face time each day.

It's far from perfect, but I hope to share some of this with the community.

Designing the setup flow was harder than I expected. In addition to the challenges of Bluetooth, customers are setting up this device in their cars. What if they don't have reception or don't want to update over cellular? There are so many uses cases and errors to consider...

I learnt that it's important to create a fixed number of pathways and to try and funnel the different flows into them. Otherwise you're creating an implementation nightmare.

If you experience a seamless setup on Google Home etc, give those designers a clap.

Onboarding was fun to design as always. We focused on teaching people how to ask for things and navigate around with voice and dial. Here's a WIP of an older prototype.

Designing for the car context requires a different approach to mobile because, as a user, you have split attention and limited abilities. To help with this, we added small pieces of logic that I hope no one ever notices, but aids them in achieving what they're trying to do.

We spent a lot of time working through navigation, especially given the small screen and challenges of touch in the car. The main questions were how you return to the now playing screen and tracklist, and discussions about global consistency, task optimization and hidden logic.

To return to the now playing screen, you press the back button 2-3 times. Cumbersome? Maybe, until you consider that you can find the button without looking, you don't need to change your hand grip or verify the result of your actions, and you can't mis-tap the screen...

But in most cases, you'll be navigated back automatically after a certain time. Unlike on mobile, users in the car might abandon a task (looking for a podcast) and then minutes later need to perform a completely different task (skipping a song).

Ok! That's a quick look at some of the process and thinking. Thanks for tuning in

If you snag yourself a device, please let me know how we can improve it!

Latest Threads Unrolled: