#56 Skyline

Inspiration

The size of the Urban Air Mobility Market (think of an air-Lyft) is expected to reach $28 Billion by 2030. Still, policy-makers, startups, and many others have yet to develop a framework for integrating this ground-breaking technology in an invisible and sustainable, yet efficient manner.

With the boom of AI, it's no longer crazy to believe that very soon these systems will be completely unmanned, which begs the question: How are we going to navigate such a chaotic sky?

What it does

SkyLine is a real-world, 1:1 simulation engine which emulates an intelligent routing system, that you'll likely see one day. It takes a 3D mesh of NYC and deploys hundreds of aircrafts, simultaneously managing multiple Collision Detection (ray casting) systems, optimal routing algorithms (Djikstra's & Bi-directional A-Star), and congestion monitoring.

SkyLine provides insights for anyone curious, into how we can actually leverage these technologies in a safe, sustainable, and human-friendly manner.

How we built it

SkyLine was built on the back of Unity with the necessary help of graph theory to tackle efficiency. We dynamically created nodes at the top of buildings then developed our own adaptive layering system to assign each aircraft to one of n planes, which served to reduce collisions by 97%.

We then implemented vertical 'columns' allowing UAMs to ascend to their designated layer, and route themselves intelligently to their destination node.

Challenges we ran into

None of our team members had ever touched C# nor the Unity engine, but unfortunately we a little time and a big idea. For the first time in working together, we had to find ways to help each other when we all had no knowledge.

Furthermore, our idea was so unheard of that we couldn't even find resources on YouTube, StackOverflow, and ChatGPT could only do so much.

Accomplishments that we're proud of

The biggest accomplishment we had was the moment when we realized that this project made no sense if it needed to be hardcoded for every city. It was sort of a lightbulb moment, but we wish we'd realized sooner than 8 hours before submission.

Still, we finished our project, creating a fully adaptive system, that given any 3D layout, can manage itself, all within a handful of hours, while managing to meet so many cool new people along the way.

What we learned

We learned early on that while we knew we had no experience with the technology, we each had a different set of skills that we could leverage uniquely. Furthermore, we learned what the values of simulations really was. Not just to provide a model of the real world, but to experiment, test, break things, and learn more about about what's possible, and how we can do things better.

What's next for Skyline

Performing data analysis to show how Skyline improves traveling experience by mitigating traffic.
Deploying this system over many cities, to explore more edge cases and create a well-rounded product.
Incorporating a better UI to allow more more data insights such as traffic slow / congestion monitoring visualization.