PinPoint

Inspiration

When I first visited University of Wisconsin Madison, I was amazed by the huge campus. There were lots of events in so many different buildings, and it was easy to get to each building with the help of google maps and buses. However, once I entered a building, I was all on my own. There was nothing but my bad sense of direction and the sparsely placed indoor maps around buildings-- UW Madison buildings are enormous, especially the lecture halls. It was not just me as well; I have friends who also talk about wandering around buildings for almost an hour looking for their TA office or a club meeting. One of our friends would constantly notice this person in a wheelchair and he always wondered how long it must've taken them to find their way around campus and especially inside the building.

What it does

Similar to Google Maps, our app finds the location of the user using networking and uses a path finding algorithm to find a way to reach their destination with ease. GPS, however does not work well, but we found a solution through the inner workings of networking.

How we built it

First we needed to find the location of the user in a building. Secondly, we needed to construct a map or layout of the building. After that, we implemented an algorithm that would produce the shortest path to the destination. When all of the computing is finished, we display the results on a web application.

Challenges we ran into

Since every framework/platform offered different advantages which some others lacked we had a difficult time choosing which one to use and this affected our efficiency as we kept switching between networks. Also, managing the team and delegating everyone doable yet meaningful challenges were difficult.

Accomplishments that we're proud of

What we learned

We learned about the inaccuracies with using bluetooth as a way to triangulate location. Because of the noise in bluetooth signals, it was impossible for us to accurately figure out the user's current location without using other sensors. As a result, we learned that we would have to also use an accelerometer along with the bluetooth beacons to accurately find the users location.

What's next for PinPoint

Our project requires many resources, such as wifi modules and lidar data. If we have the data of these two items, we can fully integrate our project and present our original idea and goal. In an article to more accurately calculate a user’s indoor positioning there are hybrid models that combine trilateration with an algorithm called dead reckoning, which measures how far a person walked using the phone’s accelerometer and gyroscope. However, it doesn’t know the initial location and after accumulation of sensor data inaccuracies, it gets inaccurate. Since it works for our simple environment, we also hope to expand it to more complex and multi story maps. And of course, we want to improve our ui to let users have better experience while utilizing our service.