Aero-Grid

A full-stack visualization of four classical AI techniques cooperating to plan and execute a multi-stop drone delivery mission across a 40x40 city grid. Naive Bayes classifies weather conditions for a pre-flight go/no-go verdict. A Genetic Algorithm solves the delivery order as a TSP variant. A* pathfinds each individual leg around buildings and no-fly zones. Q-Learning trains a tabular policy and stress-tests it under obstacle perturbation. Every algorithm step is visualized in real time — generational fitness curves, A* frontier sweeps, Q-table heatmaps. FastAPI backend exposes each module as a stateless endpoint; the city grid is the frontend's state, passed with every request.