Lindenmayer systems have been developed to model plant growth. An evolutionary algorithm can be used to simulate competitive plant evolution using L-systems. The systems themselves represent DNA, while the structures they produce represent plants. This technique gives rise to interesting yet familiar plant-like structures.
Terrain features are to a high degree shaped by hydraulic erosion. Several algorithms exist for simulating this process that vary in realism and computational complexity. This article proposes a simple and fast method that produces realistic looking results.
Artificial neural networks can be trained to perform certain tasks, but they can also be the product of simulated evolution. Evolved nervous systems are created according to a genetic blueprint. This article demonstrates squids swimming efficiently using tentacles controlled by an evolved neural network.
Voxels scenes can be rendered using layered voxel rendering by splitting the scene into image layers and pre-shading the voxels. The method is trivial to implement and surprisingly performant. This article walks through the process and demonstrates a procedural island generator using this technique.
Self-updating textures can be used to achieve impressive visual effects using the GPU. In this article I demonstrate four interactive applications of this technique, ranging from simple cellular automata to displacing vegetation and simulating waves in a body of water.
Ecosystems can be simulated to observe and understand their behaviour. In a closed ecosystem, nothing enters or leaves the system from the outside world. I simulate and explore such a system in this article, and I discuss its real world similarities and applications.
Swarm behaviour can be simulated using a few simple rules. Implementing these rules enables us to simulate flocking birds, schooling fish and swarming crowds. In this article, I walk through the process of implementing this algorithm.
Lindenmayer systems (or L-systems) can be used to produce intricate patterns by repeatedly applying production rules. Rendering these systems results in interesting fractals that can be useful in procedural generation. A method for rendering 3D Lindenmayer systems is demonstrated.
In this tutorial I walk through the process of writing a simple grid based platformer in javascript. The result is a universal platform physics engine suitable for any platformer. This implementation uses edge collisions instead of cell collisions, which allows for more compact level design.
Coherent random noise can be used to create a great variety of effects. The algorithm for generating cubic noise is explained. An interactive cubic noise generator is inclued, as well as a procedural terrain generator.
