Happy New Year!
We hope everyone enjoyed their holidays and had the chance to spend quality time with loved ones, family, and friends. Wishing you all the best for 2025! 🎉

Project Update:

Over the past two weeks, we made significant progress and discoveries that will shape the development of Hive Mind moving forward.

1. Learning from Simulation Testing

Before implementing the food component, we decided to stress-test the simulation by introducing thousands of ants. To stay above the 60fps threshold we only could simulate 200 ants. This exercise revealed the need to rethink the overall ant setup and explore performance optimizations. One of the critical findings was that simulating gravity within the tick function significantly impacts the game thread performance. We need to simulate gravity as we are using the UFloatingPawnMovement component which has no gravity simulation enabled by default. We briefly switched to the Character class and using its UCharacterMovement component which has gravity simulation enabled by default. But this did not yield the expected performance improvements (it even got worse). These results highlight the importance of revisiting our system architecture in total to ensure scalability and efficiency before adding new components like food and combat mechanics. These new and existing mechanics are not gone or of the table but they will only move to a later phase of in our development process.

2. Transitioning to Custom Environments

We also learned that the Unreal Engine landscape environment is not ideal for our simulation’s needs. To address this, we’ve started learning using Blender to create a custom-designed landscape tailored specifically for Hive Mind. The first prototype is a 5m x 7m area designed with two ant colonies in mind, ensuring symmetry to avoid any terrain-related advantages for one colony over the other. We opted for a low-poly style to prioritize performance and aesthetic simplicity.

3. Streamlining Updates

We’ve decided to skip weekly video updates for now. Producing videos is time-intensive, and since there’s currently little visually compelling content to showcase, we’re focusing our efforts on development. Instead, we’ll continue to share insights and visuals like screenshots to keep everyone in the loop.

4. Looking Ahead

Our next focus is finalizing the custom environment in Blender and testing it in Unreal Engine. Once the environment is in place, we’ll experiment to determine how many ants the simulation can handle comfortably. Performance optimization strategies like object pooling, level streaming, and advanced game mechanics will follow to make Hive Mind both immersive and efficient.

We’re excited about what lies ahead and appreciate your continued support as we work toward creating something truly special. See you in the next update!

Cheers,
The Hive Mind Team