PROJECT ID: AXK-2026 // QUANTUM MACHINE LEARNING
Axion Kitchen
Winner of the QPoland Quantum Hackathon 2026 main prize and joint winner of the Best use of VeloxQ award with the independent team stegoqubit. We turn catering orders into validated production schedules using QAOA, VQE and VeloxQ.
Project Overview
The Problem
A catering kitchen has to prepare, cook and pack main and side courses before fixed departures. A meal counts only when both courses are ready, so one missing batch can waste a whole order. The plan sets batch sizes, cookers, workers and start times on a 10-minute grid.
The Approach
One bit selects one scheduled batch program. The objective ranks missed complete meals first, then discarded food, then new batches. QAOA and VQE run on 31 to 35 data-qubit portfolios in simulation. VeloxQ solves larger QUBOs of the same model at 486, 998 and 2,022 solver bits. An independent checker validates every schedule, with CP and simulated annealing as classical controls. No quantum advantage or physical QPU result is claimed.