DrivAR
Description as a Tweet:
We're improving driving for the common good. With AR we can help everyone drive safe in a new fun way! This app is designed with the driver in mind and aims for a nearly hands-free experience for people all around the world.
Inspiration:
I'm a very nervous driver and I always go down wrong turns or make rash decisions. I wanted to make an app that could help me become a safer and more confident driver. Many navigation apps aren't user friendly and seem very clunky so I wanted to make an app that's accessible for all users and is sleek and streamlined. After some research, I decided that Augmented Reality could become my solution.
What it does:
Using augmented reality, DrivAR helps navigate roads visually. While still giving text and audio cues, DrivAR will show you the path to take to your destination in real time. Points of interest such as schools, stores, gyms, and more are displayed as you drive by so you'll always know where the nearest Chipotle is!
DrivAR also has visual and audio indicators and warnings that alert you to pedestrians, bikers, and cars that are stopping suddenly. There are also speeding warnings to avoid speeding tickets!
DrivAR can be used in multiple languages such as English, Spanish, and French with all text and audio appearing in your native language. Your preferred unit of measurement can be set as well. DrivAR supports Metric and Imperial units and will visibly and audibly tell you speed limits in either MPH or KPH.
How we built it:
This was built using the Swift programming language using Apple's Xcode IDE. Point of interest locations are queried through Radar.io. MapBox is used for directions and to help with Augmented Reality positioning through their Vision SDK.
Technologies we used:
- Swift
- Objective C
- Misc
Challenges we ran into:
Using multiple libraries with similar dependencies that required different versions was a nightmare to sort out but eventually dependency issues were resolved after a few hours. The biggest challenge however, was the ability to test the app. Since the app is a driving/navigation app, you need to drive in order to test it. Driving and coding is illegal (I assume) so I had to come up with a way to simulate it. By driving and recording my screen and keeping track of GPS and Accelerometer data, I could then use that data to simulate driving in order to test. However, due to this, the AR portion of the couldn't be tested since it requires a live camera feed. So many car trips occurred to test the AR capabilities.
Accomplishments we're proud of:
I'm proud of using speech to text and text to speech in my app, I've never worked with those technologies before so this was a great way to get into it. I've used augmented reality before, but never to this extent so being able to line up a pathway along a road is pretty cool. I'm also proud of the localization aspects of the app. I've been striving to be more inclusive with my app design and adding additional languages as well as accessibility features makes me happy.
What we've learned:
I learned about the difficulties of computer vision and augmented reality. Both of these are emerging technologies that haven't been perfected yet, so being patient when not getting the results I wanted was something I had to learn. I also learned about Radar.io, I have never heard of them before but setting up and using their API was super simple and I'm eager to experiment with it. Learning how to localize apps for different languages was also new for me.
What's next:
More languages! The goal is to have most common dialects available. I also want to continue work on the AR portion since it can sometimes glitch out, especially on corners.
Eventually, the goal is to move away from cellphones and to instead project the UI onto a car windshield where it will help drivers of all kinds.
Built with:
Swift, Xcode, Radar.io, and Mapbox
Prizes we're going for:
- Best Documentation
- Best Venture Pitch
- Best Mobile Hack
- Best Beginner Software Hack
- Most Creative Radar.io Hack