Quantum Computing Tackles Freight Logistics to Boost Fleet Income

Researchers at IonQ and Einride, led by Miguel Ángel López-Ruiz, have developed a new quantum optimisation framework to address the Shipment Selection Problem (SSP) within electric freight logistics. The framework directly tackles revenue losses stemming from the inherent unpredictability of shipment cancellations, a significant challenge for modern logistics networks. By formulating the SSP as a Mixed-Integer Quadratic Program (MIQP) and employing an enhanced Quantum Approximate Optimisation Algorithm (QAOA), the team has demonstrated improvements in operational efficiency and fleet utilisation. This end-to-end hybrid workflow, seamlessly integrating quantum simulations with existing vehicle routing software, achieves up to a 12% improvement in shipments delivered and reduces total drive distance by up to 6%, representing a key advance in operational efficiency for logistics networks.

Quantum computation optimises electric freight logistics through network-wide schedule compatibility

Shipment delivery rates improved by up to 12% when utilising this novel hybrid quantum-classical workflow, a performance gain previously unattainable with conventional classical methods. Traditional algorithms typically address shipment cancellations in isolation, treating each event as a separate optimisation problem. This approach fails to account for the complex, interconnected dependencies across multiple vehicle schedules, representing a fundamental limitation in the context of electric freight logistics where route planning and battery management are critical. The integration of Einride’s established vehicle routing software with IonQ’s quantum simulations facilitates compatibility-aware assignments for filling shipment gaps, effectively optimising the entire network rather than focusing solely on individual routes. This holistic approach allows for a more nuanced consideration of logistical constraints and opportunities.

A maximum of 130 qubits were utilised in the quantum computations to achieve a simultaneous 6% reduction in total drive distance per shipment, all while maintaining consistent operational costs. This highlights the potential for substantial efficiency gains without incurring additional expenditure, a crucial factor for logistical viability. The system’s performance is evaluated using a Schedule Compatibility Score (SCS), a metric specifically designed to quantify the degree to which new shipments integrate seamlessly with existing routes without introducing logistical conflicts or disruptions. Further assessment of the SCS demonstrated the workflow’s effectiveness in identifying assignments that minimise geographic displacement, thereby reducing the need for vehicles to deviate sharply from pre-planned paths. This is particularly beneficial in scenarios characterised by a high density of delivery locations, where even small deviations can significantly impact overall efficiency. Despite achieving improved shipment delivery rates and reduced drive distances, the hybrid approach consistently maintained operational costs, avoiding any increase in expenditure. Detailed analysis of the SCS revealed improvements across all tested instances, although the magnitude of these improvements varied depending on the complexity of the network. This suggests the potential for developing tailored optimisation strategies, adapting the quantum-classical workflow to specific logistical demands and network characteristics.

The formulation of the Shipment Selection Problem as a Mixed-Integer Quadratic Program (MIQP) is central to the framework’s success. MIQPs are known to be computationally challenging for classical solvers, particularly as the problem size increases. The quadratic nature of the program arises from the inter-gap dependencies, the way in which accepting or rejecting one shipment affects the feasibility and cost of accepting or rejecting others. By mapping this MIQP to an Ising cost Hamiltonian, the problem becomes amenable to solution using quantum annealing or, as in this case, the Quantum Approximate Optimisation Algorithm. The Iterative-QAOA algorithm employed is a variation of QAOA designed to iteratively refine solutions, leveraging the principles of quantum superposition and entanglement to explore a vast solution space more efficiently than classical algorithms. The use of 130 qubits represents a significant step towards tackling real-world logistical problems with quantum hardware, although scaling to even larger problem instances remains a key challenge.

Quantum optimisation shows initial promise for resilient freight delivery

Efficient and dynamic rescheduling following shipment cancellations is paramount for maintaining the resilience and profitability of electric freight networks, and this work offers a promising, albeit preliminary, step towards achieving that goal. While demonstrably improving shipment delivery performance, the benefits of the hybrid quantum-classical workflow are currently confined to “specific instances” as outlined in the research, leaving open the question of its broader applicability and generalisability. The research acknowledges a current lack of detailed information regarding the precise range of scenarios where these gains consistently hold, raising legitimate concerns about performance in more complex, dynamic, or unpredictable logistical environments. Factors such as real-time traffic conditions, unexpected road closures, and fluctuating demand patterns could all impact the effectiveness of the algorithm. Nevertheless, this work represents a valuable proof of concept for integrating quantum computing into real-world freight optimisation, demonstrating the potential for quantum algorithms to address complex logistical challenges, despite the current limitations to specific logistical scenarios.

Iterative-QAOA proved particularly effective as a ‘warm-start’ for classical solvers. This means that the approximate solution generated by the quantum algorithm is used as an initial starting point for a classical optimisation algorithm, allowing it to converge more quickly and efficiently. This hybrid approach leverages the strengths of both quantum and classical computing, combining the quantum algorithm’s ability to explore a large solution space with the classical algorithm’s ability to refine and optimise solutions. By considering interactions between different vehicle schedules, the framework improves overall fleet efficiency and reduces the impact of shipment cancellations. The potential for further refinement of the Iterative-QAOA algorithm, coupled with advancements in quantum hardware, could lead to even more significant improvements in logistical performance. Future research will likely focus on expanding the scope of the framework to encompass a wider range of logistical constraints and scenarios, as well as exploring the use of more advanced quantum algorithms and hardware architectures.

The research demonstrated that a quantum optimisation framework, utilising Iterative-QAOA on up to 130 qubits, can improve electric freight logistics. When used as a warm-start for existing classical solvers, the hybrid approach achieved up to a 12% increase in shipments delivered and a 6% reduction in total drive distance per shipment in specific instances. This suggests that quantum computing can offer compatibility-aware assignments that enhance fleet efficiency and mitigate the impact of shipment cancellations. The authors intend to expand the framework to include a wider range of logistical constraints and scenarios in future work.