D-Wave and Save-On-Foods Pilot Quantum-Driven Warehouse Optimization in Canada

Quantum Leap in Retail Logistics: Canada's Early Experiment

While many associate quantum computing with aerospace or pharmaceuticals, October 2020 brought a surprise development from the retail supply chain sector. Canadian grocery giant Save-On-Foods entered into a pioneering partnership with D-Wave Systems, a leading quantum computing company based in Burnaby, British Columbia. The goal: use D-Wave’s hybrid quantum-classical solvers to optimize warehouse operations — a traditionally rigid and heavily manual domain.

This partnership underscored a growing realization among grocers and third-party logistics (3PL) firms: that classical optimization tools alone are no longer sufficient to meet the pace and complexity of eCommerce-driven logistics, especially during periods of high demand such as those seen during the COVID-19 pandemic.

The Challenge: Grocery Warehousing Under Pandemic Pressure

Save-On-Foods, part of the Overwaitea Food Group, operates over 170 grocery stores across Western Canada and runs regional distribution centers (RDCs) responsible for delivering food, household essentials, and health items to stores within tight delivery windows.

In early 2020, demand spiked sharply due to pandemic buying behavior. Distribution centers faced the dual burden of:

• Rapid SKU variation as consumer demand shifted

• Congested picking aisles, increasing worker fatigue and delays

• Greater reliance on automation due to social distancing policies

Existing warehouse management systems (WMS) used heuristic-based optimizers, but failed to adapt in real time to sudden changes in order volume, labor availability, and routing disruptions.

Why Quantum? Why D-Wave?

D-Wave, known for its quantum annealing approach (as opposed to gate-based models used by IBM or Google), had been promoting use of its Leap™ hybrid quantum platform for discrete optimization problems. These include:

• Vehicle routing

• Job shop scheduling

• Bin packing

• Graph traversal (such as picking pathfinding)

In October 2020, D-Wave worked with Save-On-Foods to design and test QUBO (Quadratic Unconstrained Binary Optimization) models that represent warehouse picking problems — specifically focusing on how to optimize pick path and packaging sequences within the RDC in Langley, British Columbia.

Using D-Wave’s hybrid solvers, the teams simulated thousands of item-picking combinations per shift, constrained by:

• Worker proximity rules (to avoid aisle congestion)

• Expiry or freshness prioritization for perishables

• Real-time traffic within aisles and between zones

The results were encouraging. In several controlled simulations, D-Wave’s model generated picking paths that improved time-efficiency by 12–18% compared to the WMS baseline.

Technical Breakdown of the Use Case

1. Formulating the QUBO

The picking problem was modeled as a graph traversal problem, where each node represented a SKU location and edges represented the path cost (distance and congestion). The optimization objective was to minimize total distance and time while respecting real-world constraints like restocking delays and item priority.

D-Wave’s engineers collaborated with Save-On’s data science team to encode the problem into QUBO format, then sent the problem to the Leap platform, which routed the query through a hybrid annealing solver.

2. Post-Processing Integration

While D-Wave generated optimized pick lists and routes, Save-On used middleware to re-inject the output back into their ERP/WMS software to direct actual human or robotic picking systems.

Though it was a closed-loop simulation, the groundwork for real-time quantum-assisted routing was laid. D-Wave representatives noted that future versions of the system could be deployed in real-time operations via edge computing nodes located in RDCs.

Broader Implications for Quantum Logistics

This proof-of-concept marked several key firsts:

• First commercial quantum application in grocery warehouse operations

• First North American retail logistics firm to use a cloud-based quantum solver in a live business scenario

• A demonstration of hybrid systems (quantum + classical) outperforming traditional heuristics in retail use cases

The Save-On–D-Wave pilot also validated that you don’t need full error-corrected quantum computers to achieve operational benefit. Instead, quantum annealers and hybrid systems are already commercially useful when the problem is well-constrained and mapped effectively.

Global Comparisons: Who Else Is Exploring This?

While this was a Canadian first, similar efforts were emerging worldwide:

• Mitsubishi and Toshiba in Japan announced warehouse route optimization projects using quantum-inspired classical algorithms (simulated annealing on GPUs).

• Volkswagen and Covariant explored combining robotic picking with quantum scheduling simulations in Germany.

• Alibaba DAMO Academy initiated a project to use quantum algorithms for warehouse bin packing and inventory rebalancing in its smart warehouse in Hangzhou.

However, Save-On-Foods and D-Wave took a unique step by fully integrating a commercial quantum platform into an actual grocery RDC workflow simulation — rather than keeping it as a theoretical academic study.

Constraints and Skepticism

Despite the excitement, both parties acknowledged limitations:

• Scaling Challenges: Quantum annealers like D-Wave’s 2000Q (used in 2020) had limited capacity for very large problem sizes, necessitating smart partitioning.

• Cold Start Limitation: Hybrid solvers require solid classical approximations to function well — if the classical baseline is poor, the improvement from quantum is reduced.

• Talent Requirements: Training retail logistics staff to work with quantum modeling remains a long-term hurdle.

Still, these challenges are not insurmountable. As D-Wave expands to Advantage™ systems (released in late 2020 with 5000+ qubits), future scalability looks more promising.

What Comes Next?

Following the successful simulations, Save-On-Foods announced it would:

• Expand quantum trials to multi-warehouse coordination, examining how to split orders across different RDCs efficiently

• Integrate real-time traffic data into the QUBO models, especially during peak periods (e.g., holidays)

• Work with D-Wave to explore post-quantum security layers for order data transmission within the supply chain

D-Wave, meanwhile, launched a retail-focused QUBO template in its Leap platform, allowing other grocers and 3PLs to test similar problems using anonymized sample data.

Conclusion: Retail Logistics Finds a Quantum Advantage

The Save-On-Foods and D-Wave pilot in October 2020 was a landmark moment — not because it solved quantum computing’s grand challenges, but because it delivered measurable operational gains in a real commercial environment. In a sector often viewed as slow to adopt bleeding-edge tech, this collaboration showcased the transformative potential of quantum logistics when carefully scoped and executed.

As quantum computing matures, grocery supply chains — long the realm of tight margins and high complexity — could emerge as one of its first mainstream beneficiaries. This experiment provides a repeatable model for logistics professionals worldwide: start small, iterate often, and bridge the gap between theory and warehouse floor.