3D Nav

What it does

3D Nav is a navigation system for flying cars that encompasses the added lanes, as flying cars will have two more directions-up and down. It is able to help a flying car avoid random obstacles, such as cars, and allow the car to go from a specific starting point to a point where there is no traffic.

How we built it

We built it by having the code split up into different sections. We have a traffic generating section, where the code makes the traffic(which we referred to as the maze), and creates obstacles and starting points. It also makes the initial pathways and fake pathways(which will be where there are dead ends). We also have an AI focused section, where we have a recursive method to find the correct path, and print the coordinates so we can see what the navigator is seeing. It also prints what direction to go to achieve this path.

Challenges we ran into

We had to account for many different cases, as there can be so many when going through traffic. That complicated things for our code. Implementing the correct recursive algorithms and figuring out how to abruptly break out of them was also a challenge.

Accomplishments that we're proud of

With this type of code, there are a lot of different cases that could happen, but we were able to think of so many different cases. This was also one of our member's first time coding in Java, which was a huge accomplishment. We had some of our first experiences with AI during this, so that is a big accomplishment as well. We were also proud that the code didn't have stack overflow errors.

What we learned

We each learned different things. Lindsay learned a new language, Java, and learned to understand the different formats and how to use these ideas that could work with Python and translating it into Java. Christopher learned how to use recursion to work with AI, and how to forcefully break out of the recursion.

What's next for 3D Nav

We believe that implementing image recognition is the next step. We had random points to signify the cars so we could set the "traffic" for the AI to go through, but if we were able to use image recognition, the random points could be replaced by actual cars. This will allow 3D Nav to be used while driving in an actual environment at real time.