A client approached me with an interesting challenge: creating a VR simulator where users could shoot drones from a swarm, forcing the remaining drones to dynamically regroup and reform their original shape.

While this may seem straightforward, efficiently reorganizing a swarm after drones are removed is a complex optimization problem. Each remaining drone must be assigned a new position in the formation while minimizing unnecessary movement.

To solve this, I used the Hungarian Algorithm to optimally map the available drones to a set of desired positions. This approach minimizes total travel distance and enables the swarm to quickly reorganize into different formations, such as grids or circles, in a smooth and efficient manner.

The system is not limited to predefined layouts. By defining a set of target positions, it can generate and reorganize drones into virtually any formation shape. Below are two simple examples, a grid and a circle, but the same approach can be extended to custom patterns, logos, text, or any arbitrary arrangement.