The data you create with Automatic tells a story. But graphs and stats and numbers only provide part of the tale. That’s why we created Automatic Dashboard Labs, a new way to visualize and interact with your driving data. And lead platform developer Brendan Nee brought it to life.
Brendan’s been with us a little over two years, and if you’ve interacted with nearly anything on Automatic.com, you know his work.
Image may be NSFW.
Clik here to view.
Most people prefer to use our mobile app to keep track of their driving information, but Dashboard opens a new world of analysis and accessibility. If you’re keeping tabs of your driving over days, weeks, months, or years, it’s a powerful tool, and for hardcore data nerds, you can even download all your driving information to import it into a spreadsheet.
But Brendan saw the potential to do so much more.
“We’ve always wanted to expand Dashboard,” says Brendan. “But the design constraints meant we needed a place to do sloppy work quickly.” That’s how Dashboard Labs was born.
Automatic – like many other tech companies – allows devs to take time out of their normal gig to have fun with other projects. This “20 percent” doctrine has brought a handful of features to Automatic over the years (and much more to come), and Brendan’s work with Dashboard Labs is a perfect example of what kind of potential lies untapped in the Automatic platform.
Image may be NSFW.
Clik here to view.
One of those features is the mesmerizing Hyperlapse Generator, which stitches together images pulled from Google Street View into a video of your drive. But it wasn’t exactly easy.
“That was definitely the most tricky one,” says Brendan.
In order to pull the correct images, Brendan had to use one open source library that taps into another open source library. But stitching together disparate elements wasn’t the hard part: it was getting the camera pointed in the right direction.
Street View takes images using a 360-degree camera. If you just pull in GPS data from that library, it could point in nearly any direction. The solution was to determine the heading of the vehicle and then update the function to have it display the direction of travel. And that had to be done at every point in the trip, particularly at turns.
“We would like to be able to pause it and look around,” says Brendan. “But that would be a lot more work,” something outside developers could tackle themselves.
Image may be NSFW.
Clik here to view.
The Heat Map and Trip Line integrations were clear ways to use driving data to make visually compelling, interactive images, and all of them use Mapbox because of its flexibility, ease-of-use, and focus on design. But maps are the obvious choice and Brendan wanted to get out of his comfort zone.
“I wanted to do something that wasn’t just maps,” says Brendan. “I have a hard time not doing a map and Dashboard does a good job of graphs.” So instead he created the Carbon Calculator, a way to estimate how much CO2 you’ve put out while driving, providing a rough idea of how many trees you’d need to plant to offset it.
“We let you specify which kind of fuel you use and we already have your fuel economy data, so it wasn’t too hard to do the math,” says Brendan. “But what we really need is an API for carbon credits,” that way you could automatically, say, donate to a reputable group that plants trees based on your carbon output.
While that’s not happening yet, Brendan wanted to make something actionable – a way for you to analyze and change your behavior. After all, what good is data if you can’t do something with it? The Commute Analyzer does exactly that, telling you what times you’re most likely to be stuck in traffic and provide suggestions on when to leave to avoid being stuck in your car.
“This might be the closest thing we can implement,” says Brendan, noting that if any of these Labs programs gains traction they could turn into features. “Maybe we could pull it into Dashboard, or maybe we could send an email to alert drivers.”
Image may be NSFW.
Clik here to view.
All 8 features in Dashboard Labs didn’t take much time to develop. Work began about two months ago and Brendan says that it only took about 20 hours to deploy all the Labs functionality.
“A lot of these labs are pretty simple,” says Brendan. “Maybe 20 lines of code.” And they’re all built atop Automatic’s public API, so developers can create their own features and implementations. Just as importantly, everything in Labs is created on the client side – no server part required – and built using HTML, CSS, and Javascript; further proof that modern web browsers are awesome.
“There’s really nothing stopping anyone from doing something similar,” says Brendan. If you want to join in on the fun, hit up our Developer and API pages.
The post 20 Percent: The Making Of Automatic Dashboard Labs appeared first on Automatic Blog.