Inspiration

In large, often congested cities like Detroit, movement is cumbersome and information is elusive. Safeglass solves both problems. It presents important information from crime statistics to parking locations, and does so in an intuitive way.

What it does

SafeGlass is a cohesive platform, spanning multiple devices, that makes travel safer and more efficient by analyzing crime and population statistics, as well as regional landmarks.

The Google Glass component is the most significant part of this ecosystem, tracking the head movements of the wearer to create an augmented reality interface. It can aid a driver in finding local parking spots to ease traffic congestion, as well as detecting when a driver is feeling drowsy. It can help pedestrians steer clear of areas within a city that have higher frequency of crime, and direct them to safer locations by leveraging Amazon Alexa voice commands. All these functionalities are connected to a server that analyzes data and makes calculations in real time.

The platform isn't just an assortment of helpful apps; it's a proof of concept that Google Glass can be at the center of an ecosystem of related applications, communicating with the same server, that help reduce the complexity of city travel. As technology like Glass eventually becomes cheaper, it will become much more viable for mainstream use.

How we built it

We hosted a server on Amazon EC2 that could handle requests from multiple sources. The Google Glass activity uses the internal gyroscope and magnetic field detector to detect the wearer's orientation and location, and projects relevant information about the user's current activity (walking or driving) in the style of unobtrusive AR.

We also used Amazon's Alexa to create a voice interface with the main server, providing information about which areas and neighborhoods are safest in the user's vicinity. This data comes from Detroit Police Department crime reports, which are parsed and analyzed based on their proximity, recency, and severity (as determined by a sentiment analysis algorithm). This data is also accessible through a map visualizer that works on desktop and iOS, powered by Esri's mapping tools and based on ArcGIS data structures.

Architecture:

Challenges we ran into

We struggled to incorporate the large DPD records into our app, so we figured out a way to sift through the data and find the crime reports that were the most relevant to the user. The AR setup on the Google Glass was also difficult to calibrate and perfect, because there aren't many AR applications of Glass.

Accomplishments that we're proud of

We made a lot of progress during the 36 hours, and still managed to get a few hours of sleep. We worked on multiple pieces of hardware, and multiple operating systems and devices.

What we learned

ArcGIS data and Esri map visualization, Google Glass development, and parsing huge datasets.

What's next for SafeGlass

Enable more devices (like Apple Watch, iPad, Pebble, etc.) to work on the server, and apply data analysis and AR to solve more problems related to transportation.

Share this project:
×

Updates