quantum pathfinding without setting your laptop on fire.
in natural disasters or high-density facility emergencies, pathfinding algorithms face severe bottlenecks: road capacities dynamically collapse, human bottlenecks form rapidly, and traditional Dijkstra / A* graphs scale quadratically under real-time re-routing.
quantum heuristic search and QAOA (Quantum Approximate Optimization Algorithm) offer exciting theoretical bounds for combinatorial graph problems. the catch? real quantum computers are noisy, and simulating them on standard classical computers means your state vector doubles in memory with every single qubit you add ($2^N$ complex amplitudes).
what quanevac tackles
1. Graph Decomposition: instead of attempting to throw an entire city graph into a single quantum circuit, Quanevac divides the evacuation grid into regional subgraphs, solving local choke points before routing globally.
2. Bounding Memory Usage: classical simulation requires strict state-space pruning so evaluation circuits don't exceed 16GB of local memory.
3. The Reality Check: this project taught me that mathematical elegance on paper means nothing if you don't account for the physical memory constraints of the machine running it.