For years, Germany has been searching for various ways to better manage its chaotic rail schedules—from digital signaling to new apps. Now, an unexpected ally may be on the horizon: quantum physics.
CNMO has learned from German media that a research team is exploring how quantum computers can aid in real-time rescheduling, or adjusting train schedules while they're running. They're using the Baltimore light rail system as an example. This system is unique because trains share the tracks with cars in the city center, and random interference is commonplace.
The research is unique in that the researchers exploit not the computational precision of quantum machines but their noise. The NISQ device ("Noisy Intermediate-Scale Quantum") offers not a perfect solution, but many slightly different ones. What might seem like a flaw at first glance is surprisingly well-suited to a system where chaotic interference is inherently inevitable. The researchers, publishing their findings in Nature, show that the deviations generated by the quantum computer reflect the real-world delay patterns observed in everyday life.
The tests were conducted on two platforms: a D-Wave quantum annealer with over 5,000 qubits, capable of calculating thousands of schedule variations in seconds; and a gate-based computer from IonQ, using a different algorithm (QAOA). The results were similar to those measured in Baltimore: many trains arrived nearly on time, but certain trajectories exhibited a so-called "long right tail." Statisticians use this term to describe a frequency distribution: it typically shows small delays, but also includes rare, significantly longer extremes. This is a pattern familiar to every commuter in the real world.
However, the current financial costs remain substantial: while a classical computer can solve a (no interference) schedule model in milliseconds, the cost of a real-time quantum experiment ranges from approximately $1,000 to $67,000, depending on the platform. This is clearly unsuitable for practical applications. However, the experiment demonstrates that quantum noise is not just a technical issue; it can also serve as a tool to better model the unpredictability of rail networks.
Looking ahead, the researchers are cautious, emphasizing that current quantum computers are too small and expensive. They see potential in a hybrid approach: classical computers manage the stable parts of the network, while quantum processors simulate the perturbed parts, which fluctuate under external influences. This approach could, over the long term, provide a realistic picture of complex transportation networks. As next steps, they propose improving error correction, developing more efficient algorithms, and transferring the technology to other systems, such as subway or bus networks.
