Genetic Algorithm Evolution Simulator

Genetic algorithms search for good solutions the way evolution does: keep a population of candidate answers, score each one by how well it performs, let the fittest reproduce, and repeat. This simulator makes that loop visible. A population of agents starts each generation at the same line and drives across a 3D landscape scattered with obstacles toward a fixed goal, each agent steering according to its own chromosome — a short sequence of turn genes. Agents that get close to the goal (or reach it) score well; agents that crash into an obstacle are penalized. At the end of every generation the weakest agents are discarded, the fittest survivors crossbreed and mutate to produce the next population, and the cycle repeats — watch the fitness graph climb and the swarm's paths tighten into a clean route as it learns.