Generation: 0
Best Score: 0
This demo uses a simple genetic algorithm to evolve a population of birds. Each bird has a set of weights that decide when it jumps. When all birds die a new generation is spawned based on the best performers with random mutation. Watch as the birds get progressively better at navigating through the pipes.
The maze is generated on the client using a depth first search algorithm. Cells are carved out by recursively visiting neighbours in random order.
Click to set the start and end points then add walls. Running the visualiser animates the A* algorithm searching for the shortest path. Check the diagonal option to allow diagonal movement.