Maersk and QC Ware Launch Quantum Pilot to Optimize Inland Haulage Logistics
April 5, 2021
Maersk's Inland Logistics Challenge
As the world’s largest container shipping company, Maersk operates not only at sea but also across vast inland haulage networks. These cover tens of thousands of trucking routes connecting ports to final destinations. The complexity of optimizing this network in real-time has become a critical challenge as supply chain disruptions, carbon regulations, and customer demands grow more volatile.
In response, Maersk launched a pilot with QC Ware to explore whether near-term quantum machine learning (QML) tools could enhance dynamic routing and fleet deployment decisions beyond what current algorithms offer.
Project Goals: Sustainable Optimization at Scale
The core objective of the April 2021 pilot was to:
Reduce carbon emissions by minimizing truck mileage and idling.
Improve cost efficiency via better asset allocation and load balancing.
Enhance responsiveness to port congestion and weather-related disruptions.
The pilot focused on logistics hubs in Germany, Belgium, and the Netherlands, targeting major intermodal corridors where port throughput is high and traffic conditions are unpredictable.
QC Ware’s Role: Quantum-Ready Algorithms for Trucking Logistics
QC Ware, a Palo Alto-based quantum software firm, brought to the table its Forge platform, which provides access to hybrid quantum-classical algorithms that can run on both simulators and real quantum hardware.
Key technical strategies included:
QML models trained on port data, route histories, and weather feeds.
Quantum-enhanced k-means clustering to group similar transport tasks.
Variational Quantum Circuits (VQC) used to explore optimized route groupings under constraints.
QC Ware engineers worked with Maersk’s data science and transport planning teams to translate route planning problems into quantum-classical representations suitable for hybrid computing workflows.
Simulation Environments and Constraints
The Maersk-QC Ware team ran simulations on synthetic and anonymized datasets mimicking:
Daily truck dispatch volumes from Rotterdam, Antwerp, and Bremerhaven.
Real-time traffic variability and incident delays.
Emission thresholds and delivery time window constraints.
A key challenge involved encoding vehicle routing problems (VRPs) into a format solvable by Noisy Intermediate-Scale Quantum (NISQ) devices, given current hardware limitations.
Initial Findings from the April 2021 Pilot
While full deployment on live fleets was not attempted during the pilot, initial simulation results showed:
5–8% reduction in total travel distance compared to Maersk’s classical baseline.
Improved fleet utilization efficiency in multi-stop trip scheduling.
Early signs that quantum-enhanced clustering led to more robust grouping of delivery requests under uncertainty.
The pilot also highlighted the importance of hybrid architecture, where classical preprocessing and post-processing are paired with quantum optimization cores.
Environmental and Strategic Implications
For Maersk, this pilot aligned with broader environmental and digital transformation goals:
The company has pledged to become net-zero by 2040 and seeks technologies that cut emissions without compromising service levels.
Maersk’s growing inland business demands fine-grained optimization tools as more operations move from ocean-centric to door-to-door logistics.
QC Ware’s quantum-enhanced models promise incremental improvements today and significant gains as quantum hardware matures.
Industry Context: Quantum in Freight and Route Optimization
This pilot adds to growing interest in quantum applications in transport logistics:
DHL and Terra Quantum (2022) tested quantum-inspired rerouting.
DB Schenker explored quantum-secure communications in trucking networks.
Kuehne+Nagel began investigating quantum-enhanced demand forecasting.
Maersk’s early involvement positions it to lead in the long-term shift toward quantum-optimized supply chains, especially as quantum computers become more accessible via cloud platforms.
Talent and Infrastructure Considerations
Maersk’s data science team acknowledged a steep learning curve in working with quantum tools. To address this:
QC Ware provided training and co-development sessions.
The pilot included developing explainability layers to help planners interpret QML suggestions.
QC Ware’s Forge was run in the cloud, minimizing infrastructure friction for Maersk.
Next Steps and Roadmap
Following the April pilot, Maersk indicated interest in:
Extending the quantum models to rail and barge segments in Europe.
Testing quantum optimization for container repositioning problems.
Exploring integration into Maersk’s TradeLens blockchain-based visibility platform.
A follow-up proof-of-concept phase is scheduled for 2022, contingent on improvements in quantum simulation speed and hardware access.
Conclusion: Quantum Trucking Optimization Enters the Real World
This pilot marks a quiet but important milestone: quantum machine learning is beginning to address real-world operational logistics challenges at the enterprise level.
For Maersk, it represents an initial step toward a future where quantum-enabled decision engines might operate in tandem with classical analytics to drive precision, sustainability, and competitiveness in global logistics.