IQM and Deutsche Bahn Test Quantum Scheduling on Real Rail Data

IQM and Deutsche Bahn have run a quantum optimization experiment on an actual railway schedule. The dataset covered 190 trips across five German cities, working out to around 98,500 possible scheduling cycles. The result is a working hybrid method, as opposed to being a demonstration that quantum computing beats the tools DB already uses.

How the hybrid quantum scheduling system works

According to the July 20 press release, the setup splits the job. A classical system holds the full-scale problem and carves off smaller pieces, which a quantum processor tackles using QAOA, a well-known algorithm for optimization on today’s noisy machines. The classical side then reassembles the answers. The full pipeline, from formulating the problem to producing a usable schedule, ran end to end on IQM hardware.

The most interesting finding, described in IQM’s technical whitepaper, is about scaling. The researchers reported a statistically significant link between the size of the subproblem the quantum processor could handle and the quality of the final solution. If that holds, the same framework should improve as quantum machines get bigger, without having to be rebuilt.

What the railway experiment proves

The authors described the results as feasible and good quality. The authors don’t claim they beat classical optimization, which is the comparison that would be important to a rail operator. The quantum component also does a small share of the work, so this is hybrid computing with a quantum piece inside it, and not a quantum solution to scheduling.

IQM Chief Scientist Inés de Vega said the work shows quantum “delivers value now,” and DB’s Manfred Rieck called it “another step toward quantum advantage.” Indeed, as Rieck explained:

“Quantum computing is not going away.”

What gives the work more weight than a vendor demonstration is Deutsche Bahn’s involvement and the use of genuine operational data. The companies are also upfront that this covered planning under stable conditions, not the harder problem of rerouting trains during live disruptions.