Dynamic Pathfinding

I implemented a flow field based method called continuum crowds, which helped reduce the impact of agent count on performance. A specific advantage of this technique is also that it has agent avoidance and works well with slopes in terrain.

I implemented multiple performance optimizations in order for the pathfinding to efficiently respond to changes in the world. A custom job system was used to generate the pathfinding grid and flow field in parallel.

I had to work with the PS4 SDK and make sure my code adhered to the TRC. Which meant I had to ensure multi threading compatibility with the PS4.

Custom Engine

Bit-Buster was developed using our own custom-made C++ engine called Voxagine. Voxagine was build with a custom editor which exposed our entity component system to add behavior to our world. The editor allowed for rapid iteration and testing.

As the AI programmer I was responsible for working together with designers to implement all the enemy behaviors. By using Voxagines entity component system, I could split enemy behavior into components, which allowed for reusable behavior that could be applied to multiple enemies.

Large Multidisciplinary Team

We spend 8 weeks on building the engine after which artist and designers were added to the team to work on the final game for another 8 weeks. During these last 8 weeks I became the programming lead.

As a lead I contributed to the high level planning and organized the planning for programming. Most importantly I was responsible for keeping the vision among programmers and communicating the needs and feedback of the programmers to the team. This also resulted in me giving multiple presentations on how to use the engine and established pipelines to other disciplines. This meant they could start producing assets and gameplay that adhered to the engine restrictions.