Quantum Computing Tackles Real-World Logistics, Cutting Costs and Carbon Emissions

Researchers tackled the complex challenge of multi-objective supply chain logistics, formulating a real-world problem as a Quadratic Unconstrained Binary Optimisation Problem to simultaneously minimise cost, emissions, and delivery time, alongside maintaining equitable supplier workshare. Raoul Heese from NTT Data Germany, Timothée Leleu and Sam Reifenstein from NTT Research, alongside Christian Nietner and Yoshihisa Yamamoto, present two novel hybrid quantum-classical solvers, an informed tree search (IQTS) and a modular bilevel framework (HBS), that integrate quantum computation with established classical heuristics. This work is significant as it demonstrates a viable methodology for mapping intricate logistics problems onto current quantum hardware, specifically the Aria-1 platform, and achieving high-quality, Pareto-optimal solutions with potential for substantial improvements in supply chain efficiency and sustainability.

This work addresses the inherent complexities of modern supply chains, specifically the need to simultaneously minimise cost, carbon emissions, and delivery time, whilst adhering to predefined supplier workshare targets.

The research successfully maps a real-world logistics problem, originating from Airbus, onto emerging quantum hardware, paving the way for more efficient and sustainable operations. IQTS leverages problem-specific knowledge to accelerate convergence, integrating the Quantum Approximate Optimisation Algorithm (QAOA) within an explorative tree decomposition.
HBS adopts a modular architecture, combining QAOA, Chaotic Amplitude Control with Momentum (CACm), Iterative Belief Propagation (IBP), and Dynamic Anisotropic Smoothing (DAS) to achieve computational speed and scalability. These solvers were designed to iteratively construct, repair, and refine solutions, ultimately improving both solution quality and feasibility.

The study successfully implemented this methodology using IonQ’s Aria-1 hardware, a state-of-the-art trapped-ion quantum computer. This implementation generated Pareto-optimal solution candidates for a real-world logistics network from Airbus, demonstrating the potential of quantum-classical approaches to tackle complex optimisation challenges.

The ability to generate Pareto-optimal solutions signifies a significant advancement, allowing decision-makers to explore a range of trade-offs between competing objectives. This research offers a pathway towards optimising complex supply chains, potentially leading to reduced costs, lower carbon emissions, and more efficient delivery times for businesses and consumers.

Hybrid quantum-classical solvers for multi-objective Airbus supply chain optimisation

IonQ’s Aria-1, a state-of-the-art trapped-ion quantum computer, served as the core hardware for implementing a novel methodology to address complex logistics optimisation problems. Realistic constraints integral to the problem included part dependencies, the utilisation of double sourcing, and multimodal transport options. To solve this QUBO, two hybrid quantum-classical solvers were developed: an Informed Quantum Tree Search (IQTS) and a Modular Bilevel Framework (HBS).

The IQTS method leverages problem-specific knowledge to accelerate convergence, integrating the Quantum Approximate Optimisation Algorithm (QAOA) within an explorative tree decomposition. Conversely, the HBS framework adopts a modular architecture, combining quantum techniques such as QAOA and Chaotic Amplitude Control with Momentum (CACm) with classical methods like Iterative Belief Propagation (IBP) and Dynamic Anisotropic Smoothing (DAS) in a bilevel optimisation structure.

Specialised tools, Informed Solution Generator (ISG), Informed Solution Fixer (ISF), and Informed Solution Improver (ISI), were designed to iteratively construct, repair, and refine solutions, thereby enhancing both solution quality and feasibility. These tools exploit the inherent structure of the logistics problem to guide the optimisation process. The study successfully implemented this methodology using the Aria-1 hardware, generating Pareto-optimal solution candidates for the Airbus logistics network, demonstrating a pathway towards optimising complex supply chains.

Airbus logistics network optimisation using hybrid quantum-classical algorithms

Researchers successfully mapped a real-world logistics network from Airbus onto IonQ’s Aria-1 hardware, a state-of-the-art trapped-ion quantum computer, and generated Pareto-optimal solution candidates. The QUBO model integrated realistic constraints including part dependencies, double sourcing, and multimodal transport, reflecting the complexities of modern supply chains.

Two hybrid quantum-classical solvers were proposed: a structure-aware informed tree search (IQTS) and a modular bilevel framework (HBS), each combining quantum subroutines with classical heuristics to enhance performance. These solvers leveraged specialized tools, Informed Solution Generator, Informed Solution Fixer, and Informed Solution Improver, to iteratively construct, repair, and refine solutions, improving both quality and feasibility.

The implementation of these solvers on the Aria-1 hardware yielded high-quality, Pareto-optimal solutions for the Airbus logistics network. This work focused on scalability and modularity, suggesting potential applicability to a broader range of complex real-world optimization problems and the use of special-purpose Ising machines.

The developed QUBO model optimised four Key Performance Indicators: carbon dioxide emissions, costs, time, and supplier target workshare fulfillment, addressing a comprehensive set of logistical requirements. The research highlights a pathway towards optimising complex supply chains, potentially leading to reduced costs, lower carbon emissions, and more efficient delivery times for businesses and consumers. This methodology offers a promising approach for tackling NP-hard combinatorial problems that are notoriously difficult to solve at scale using conventional methods.

Airbus supply chain optimisation via hybrid quantum-classical computation

Researchers have demonstrated a methodology for solving complex, multi-objective logistics optimization problems using a combination of quantum computing and classical algorithms. Experimental results obtained using IonQ’s Aria-1 trapped-ion quantum computer generated Pareto-optimal solution candidates for the Airbus logistics network. This achievement signifies a pathway towards optimising complex supply chains, potentially leading to reduced costs, lower carbon emissions, and more efficient delivery times for businesses and consumers.

The study acknowledges the limitations inherent in current Noisy Intermediate-Scale Quantum (NISQ) hardware, requiring hybrid approaches to manage computational complexity. Future research may focus on scaling these methodologies to even larger and more intricate logistics networks, as well as exploring advanced quantum algorithms to further enhance solution quality and computational efficiency.