Singapore’s PSA International Pilots Quantum Algorithms for Container Port Optimization
August 22, 2019
World’s Busiest Port Tests Quantum Optimization for Container Management
PSA International, the global port operator headquartered in Singapore, made waves in August 2019 by becoming one of the first logistics organizations in Southeast Asia to pilot quantum computing in real-world operations. Partnering with IBM Research, the initiative was focused on exploring quantum-enhanced optimization models for berth scheduling and container yard management at the Port of Singapore—the world’s second-busiest port by container volume.
This pilot reflected Singapore’s broader national strategy to lead in quantum and logistics innovation under its Smart Nation initiative. For PSA, whose port handles more than 36 million TEUs (twenty-foot equivalent units) annually, marginal gains in operations could result in massive global efficiency.
The Complexity of Mega-Port Operations
Managing container ports of this scale involves solving complex NP-hard problems on a daily basis. These include:
Berth allocation: Assigning ships to dock spaces at the right time.
Yard stacking: Efficiently placing containers to reduce re-handling.
Crane scheduling: Minimizing idle times and conflicts between gantry cranes.
Vehicle dispatching: Coordinating autonomous trucks and human drivers.
Each decision point affects the others. A small inefficiency in yard stacking can cause cascading delays in ship loading and departures. Existing systems, while robust, rely heavily on heuristics and AI-assisted models that may not scale effectively under pressure or uncertainty.
Quantum computing, particularly gate-based quantum systems like those developed by IBM, offer a new paradigm for handling such interdependent variables in real time.
Quantum Optimization Meets Port Scheduling
In the PSA–IBM pilot, IBM Q researchers used quantum approximate optimization algorithms (QAOA) and variational quantum eigensolvers (VQE) on simulated workloads based on historical port data from the Tuas and Pasir Panjang terminals. While the quantum processors used were still relatively small (under 30 qubits), the goal was to test hybrid quantum-classical workflows.
The pilot used a “quantum-enabled twin” of a simplified container yard, where variables like container pickup priority, berth ETA, and crane availability were modeled using Hamiltonians optimized on IBM’s 16-qubit Melbourne device via the IBM Q cloud.
Key findings included:
Up to 15% reduction in predicted gantry crane idle time in simulated runs compared to classical scheduling baselines.
Faster convergence times in multi-objective scheduling scenarios where trade-offs between energy, time, and priority are essential.
Improved adaptability to late-arriving vessels, where classical algorithms typically require manual overrides or full recalculations.
Government Backing and Talent Pipeline
The pilot was jointly supported by Singapore’s Agency for Science, Technology and Research (A*STAR) and the National Research Foundation. It fits into a larger roadmap under Singapore’s Quantum Engineering Programme (QEP), which was launched in 2018 with S$25 million in funding to support real-world quantum applications.
Additionally, NUS (National University of Singapore) and NTU (Nanyang Technological University) partnered with PSA and IBM to create joint quantum training programs, enabling port logistics engineers to begin understanding quantum optimization modeling.
Professor José Ignacio Latorre, a quantum computing expert and director of Singapore’s Centre for Quantum Technologies, commented in an August 23 interview:
“Quantum computing is not just a theoretical field anymore. This pilot at the Port of Singapore demonstrates quantum’s capacity to generate practical scheduling insights in highly congested operational environments.”
Real-World Logistics Gains on the Horizon
While PSA emphasized that the system remains experimental, the use case holds strong real-world relevance. For example, during seasonal surges or when adverse weather impacts shipping lanes, being able to recompute optimal container placement or berth assignments in near real time can prevent days of delay.
This is especially critical as PSA expands its Tuas Mega Port, which is expected to handle 65 million TEUs per year by 2040. As port capacity increases, so too will the computational burden of managing operations efficiently.
Challenges and Next Steps
Despite the excitement, challenges remain:
Hardware limitations: Current IBM Q systems are still noise-prone and offer limited qubit fidelity, restricting model complexity.
Skill gaps: Port engineers require specialized training to leverage quantum optimization models effectively.
Integration: Existing logistics platforms, like PSA’s Portnet, are built on classical architectures and require careful hybridization to interact with quantum systems.
Still, the pilot opened doors to broader experimentation. IBM and PSA have announced plans to expand the pilot to include autonomous vehicle dispatching models and real-time crane reallocation algorithms.
Quantum Port Management as a Future Standard?
If successful at scale, quantum-enhanced decision-making could become a staple in next-generation port operating systems. As global trade intensifies and smart port initiatives rise globally—from Rotterdam to Shanghai to Long Beach—Singapore’s pilot could serve as a blueprint.
Notably, the Port of Rotterdam also announced a quantum research partnership with QuSoft in the Netherlands in the same month, signaling a broader European push into quantum logistics.
Conclusion:
PSA International’s collaboration with IBM marks a significant milestone in applying quantum computing to operational logistics at massive scales. While the hardware and skills ecosystem are still evolving, early signs point to real value in hybrid quantum-classical scheduling for container ports. With ports serving as global trade arteries, quantum’s entry into their nerve centers could be one of the most impactful logistics transformations of the coming decade.