Since the release of Pokemon Go, there has been an increase in the number of pedestrian injuries in the U.S. A major factor is that people are spending more time looking down at their phone and paying less attention to their surroundings, leading to traffic accidents. Similarly, when navigating a person must constantly look down to check his or her phone, breaking focus and increasing the chance of an accident when driving or crossing a street. TravelAR can mitigate this problem and make it easier to check the route one is walking or driving without diverting his or her eyes from their surroundings.
What it does
TravelAR aids navigation by providing a minimap of the user's current position and showing the route to take on the map. Additionally, an arrow is at the top of the user's vision, showing where to move next. These features are displayed in augmented reality, allowing the user to pay attention to their surroundings while still being able to navigate with minimal distraction.
How we built it
- Google Cardboard
- Leap Motion
- Google Maps API
Our project was built using Unity, which allowed us to interface with all of the technologies listed above. We used google cardboard to display the mini-map, arrow and the menu. We sent leap motion sensor data to a server which we then queried from the phone to display where the users' hands were as well as determine which button on the menu they were touching. VuForia was used to display the hands on the screen. The Google Maps API allowed us to dynamically re-calculate the map displayed on the user's google cardboard.
Challenges we ran into
A big issue we ran into was the direction in which the arrow would display. We attempted to use gyroscope data, but it was too inaccurate for the arrow to actually seem as though it was pointing in the correct direction. To fix this we used compass data instead as well as clever use of trigonometry.
Merging of Vuforia and Google-cardboard and integration into Unity. We attempted to merge the two technologies together, but eventually decided to switch from head-tracking to compass data.
Accomplishments that we're proud of
The minimap constantly updates the user's position and is usually accurate within 10 meters. It is very similar to the minimaps that are seen in videogames and is useful for navigation without being too distracting.
The ability for us to determine, based on leap motion data which menu item the user is clicking.
What we learned
- Most of the team was unfamiliar with Unity and picked up a lot of its core concepts as we built this project.
- We learned how to integrate augmented reality, including Leap Motion, into Unity
What's next for TravelAR
- Voice command navigation (be able to say "nagivate to X" and receive directions
- Expand directions to other form of transportation, such as driving, biking, etc.