Osaka Urban Area Last-Mile Delivery Optimization Challenge 2026
openOverview
Develop an AI and quantum hybrid agent that reduces total delivery distance while adhering to capacity constraints and time windows for 16 delivery destinations and 4 vehicles modeling the Osaka urban area. In addition to delivery quality, the QUBO formulation and verifiable quantum circuits and measurement trails will be evaluated.
Background
At Kansai Logistics Co., route creation for same-day delivery in urban areas relies on the experience and manual work of the staff in charge. With demand volume and designated time slots changing daily, it is necessary to compare reproducible optimization methods and the practical applicability of quantum-classical hybrid technologies under the same conditions.
Business pain point
Manual vehicle dispatching requires long planning times, making it difficult to minimize driving distance while simultaneously satisfying the load capacity and delivery time windows for each vehicle. Furthermore, if the use of quantum technology is evaluated based solely on self-reporting, it is impossible to determine whether the results can actually be reproduced.
The task
The submitted ZIP must include manifest.json and the agent's entry point (standard is agent.py). During evaluation, QVArena will generate an official problem.json. The agent must read problem.json and output result.json containing the delivery routes, qubo.json containing the QUBO formulation, and quantum.json containing the circuit and measurement trails. Create a feasible solution that delivers to all 16 locations without duplication and satisfies vehicle capacity and time windows.