Summary of "Webinar: Quantum Meets Logistics: A Real-World Routing Case Study"

Executive summary

This webinar presented a real-world proof-of-concept routing system developed for Air Transit (a Spanish logistics company) under the Basque-funded KU-for-Real consortium. The project used D‑Wave’s hybrid quantum/classical solver (Constraint Quadratic Model, CQM) to solve constrained vehicle routing problems with industrial constraints and real road distances, producing operationally usable routes that minimize delivery cost while honoring strict business constraints (priority/time windows, 2D package/truck capacities, driver working hours, service times).

What the webinar presented

• A PoC routing system for Air Transit built by the KU‑for‑Real consortium (partners: Air Transit, Technalia, Multiverse, university partner(s), local partner recorded as Erharicat/Byaricat).
• Technical backbone: D‑Wave hybrid quantum/classical solver using CQM to express complex constrained routing problems.
• Operational goal: generate routes that are usable in production—minimizing cost while meeting business constraints such as priority clients, time windows, multi-deliveries, dimensional capacities, service times and driver hours.

Key business and operational takeaways

• Operational realism matters: academic VRP/TSP formulations often omit constraints essential to real logistics. This project prioritized realistic constraints (time windows, real road travel times, service times, 2D dimensions, multiple deliveries per client, driver hours).
• Iterative decomposition is practical: breaking the full problem into sub-routes (priority, “to the depot”, no‑priority), solving each, then merging/post-processing enables the hybrid solver to handle industrial instances.
• Solver selection rationale: CQM was chosen because it is flexible for expressing complex constraints. Binary/quadratic encodings (BQM) were less suitable here; nonlinear solver testing is planned for future work.
• Domain + solver expertise is critical: substantial upfront effort is required to translate business rules into a CQM formulation and to tune and test across scenarios.

Frameworks, processes and playbooks

Iterative sub-route decomposition playbook

1. Identify the most restrictive constraints (e.g., priority/time-window clients).
2. Build a priority sub-route (depot → priority clients).
3. Build a “to‑depot” sub-route and the remaining no‑priority routes.
4. Merge sub‑routes into complete routes and post-process.

Instance reduction / preprocessing

• Priority‑centric circle intersection:
  • Compute circles around the depot and around a priority client (radius = travel cost + offset).
  • Include nodes in the circle intersection into a reduced sub‑instance for the solver to limit problem size.

No‑priority instance reduction

• Farthest-node seeding:
  • Seed sub‑instances from farthest clients using depot distance + surplus.
  • Solve sub‑instances iteratively to partition the no‑priority set quickly.

Postprocessing (“friendly routes”)

• Reassign nearby visits into neighboring routes, when capacity and time allow, to improve driver ergonomics and respect working hours (reduce multi‑stop friction).

Solver selection decision process

• Use CQM for complex constraints.
• Consider BQM or classical tools (OR‑Tools, Gurobi) for simpler problems.
• Evaluate nonlinear solvers next as future work.

Hybrid runtime control

• Tune hybrid solver time‑limit parameters to trade runtime versus solution quality for larger instances—this is an operational knob to meet SLA requirements.

Key metrics, KPIs and operational targets

• Instance sizes demonstrated: 14, 22, 28, and 44 deliveries (real PoC: 44 orders arriving at Palma de Mallorca port/warehouse).
• Vehicles: up to 5 trucks used in tested instances; each truck can perform multiple routes subject to driver working hours.
• Tracked KPIs:
  • Number of priority deliveries served within required times
  • Total route cost
  • Total weight and total dimensions per route
  • Compliance with driver working hours
  • Truck capacity utilization
• Runtime expectations: industrial target is routes generated in seconds (example target: 5–10 seconds). Hybrid solver time limits were used to meet operational runtime constraints.
• Comparative performance: CQM hybrid results were comparable to Google OR‑Tools and other classical solvers (Gurobi); in some cases matched solution quality and achieved faster runtimes.

Concrete examples / case study details

• Palma de Mallorca PoC:
  • 44 orders delivered from port to warehouse and then routed to 44 client locations across Palma.
  • Travel times used real street routing and included per‑stop service times.
  • The solver produced operational routes and visual route maps.
• Illustrative examples used in the webinar:
  • Single priority client with a tight time window.
  • Two incompatible priority clients that cannot be served in the same route.
  • A multi‑delivery client whose deliveries may be split across routes.
  • These scenarios influenced sub‑route choices and friendly-route postprocessing.
• Demo:
  • Live runs on D‑Wave hybrid hardware were shown, including generated route files and per‑route cost/weight/dimension summaries.

Actionable recommendations (operational / product)

• When moving from academic VRP to production logistics, begin by identifying the most restrictive constraints (e.g., priority/time windows) and decompose the problem around them.
• Apply targeted preprocessing to reduce instance size before calling expensive solvers (priority circle intersection, farthest-node seeding for large no‑priority sets).
• Add a lightweight postprocessing pass (“friendly routes”) to improve driver ergonomics and reduce multi‑stop friction.
• Set and tune a time limit on hybrid solvers as an operational control to meet SLA requirements for route generation.
• Benchmark quantum/hybrid solutions against established classical solvers (OR‑Tools, Gurobi) for both solution quality and runtime—treat runtime as a critical KPI for production adoption.
• Invest in initial formulation work and cross‑functional teams (logistics domain experts + quantum/hybrid specialists). Expect formulation to be the largest upfront effort.
• For dynamic events (accidents, floods): the system can handle updated inputs, but real‑time reoptimization requires a workflow to update inputs and re-run the solver—incorporate this into operations if real‑time responsiveness is required.

Limitations and practical notes

• Dynamic, real‑time route changes (e.g., traffic incidents) were not integrated into the PoC automatic pipeline; operators must supply new inputs and re-run the solver.
• Scaling considerations:
  • The hybrid solver supports runtime caps.
  • Real logistics operations typically break daily work into many smaller route problems rather than solving a single thousands‑node instance.
  • The team has not yet solved multi‑thousand node instances in this project.
• Significant upfront formulation and testing effort is required; subsequent adaptations become progressively easier after the first problem.

Evidence of business value

• Client feedback (Air Transit) was positive: generated routes were operationally acceptable and “quite optimal” per the company’s assessment.
• Comparative tests often showed comparable or better runtimes versus classical solvers for similar solution quality—runtime advantage is a key operational benefit.

Publications, artifacts and next steps

• Open access paper in Scientific Reports describing the algorithm and PoC.
• Follow‑up arXiv paper extending to time windows and simultaneous pickup & delivery.
• Accepted paper at the Quantum Computing & Engineering (QCE) conference.
• Demo code and instance files were used in the webinar; authors offer to share papers and instances by contact.

Presenters and sources

• Presenter: Anica / Anico (transcript alternates names; presenter responsible for the solver and PoC).
• Consortium / organizations: Air Transit (customer), Technalia, Multiverse, Mercedes (university co-author name ambiguous in transcript), Erharicat / Byaricat (local partner), D‑Wave (hybrid solver and webinar host).
• Publications: Scientific Reports paper (open access), arXiv extension, QCE conference acceptance.

If needed, the presenter can also: - Extract concrete algorithm pseudocode or workflow steps for operational implementation. - Produce a short checklist for integrating a hybrid quantum routing PoC into an existing logistics IT stack.

Category

Business

---

Share this summary

Share on X/Twitter Share on Facebook Share on LinkedIn Share on Reddit Share on WhatsApp Share on Telegram

---

Is the summary off?

If you think the summary is inaccurate, you can reprocess it with the latest model.

Reprocess summary

Video