Loading an aircraft is not equivalent to filling a container. Cargo has weight, the aircraft has balance and structural limits, and placement decisions can affect what remains possible elsewhere in the hold. A load that uses space efficiently may still violate a constraint that makes it unusable.
Those interactions make aircraft cargo loading a worthwhile setting for advanced optimization research. Through the NQCC SparQ program, 4colors led work with Airbus, ORCA Computing and the National Quantum Computing Centre to examine how hybrid classical and quantum algorithms could address parts of this difficult problem. DNV participated as an observer.
The project moved beyond a circuit proposal: the consortium implemented and tested algorithms on real gate-based and photonic quantum hardware, and compared the approaches with classical baselines. That is an experimental result. It is not, by itself, a claim of proven quantum advantage in aircraft operations.
A load is feasible only when several conditions agree
Consider two pieces of cargo with the same footprint but different weights. They are not interchangeable if one placement shifts the center of gravity beyond an acceptable range or places too much weight on part of the structure. Loading order and handling constraints can reduce the options further. Every decision changes the space and weight distribution available for the next one.
Mixed-integer optimization is a natural way to represent many of these choices, though realistic instances can create demanding search spaces.
In mathematical terms, selecting and positioning cargo couples discrete assignment choices with capacity, geometry, weight and balance requirements. A good formulation must distinguish hard physical restrictions from objectives such as carrying more cargo or reducing turnaround effort. An attractive objective value cannot compensate for an invalid load.
Mixed-integer optimization is a natural way to represent many of these choices, though realistic instances can create demanding search spaces. The question is how to exploit their structure without discarding the constraints that make the answer useful.
Where a quantum routine might fit
Hybrid computation does not require the whole aircraft-loading problem to fit directly on a quantum processor. A classical algorithm can manage the overall model, construct feasible candidates and evaluate constraints, while a quantum routine is investigated for a selected computational subtask.
The distinction matters particularly because the hardware platforms differ. Gate-based systems execute sequences of quantum operations; photonic platforms manipulate information carried by light. Their native operations, connectivity and experimental constraints are not interchangeable. A method that works on one platform may need a different representation on the other.
A classical algorithm can manage the overall model, construct feasible candidates and evaluate constraints, while a quantum routine is investigated for a selected computational subtask.
The consortium examined these questions through implemented experiments and classical comparison. The existence of experiments on both hardware families is useful evidence of technical breadth. Whether either offers an advantage for a particular scale of cargo-loading problem requires measured, comparable performance under clearly specified conditions.
What the work adds to the field
This project connects an industrially recognizable optimization problem with experiments on more than one quantum hardware approach. It also reinforces a disciplined way of assessing quantum optimization: formulate the real constraints first, use a capable classical reference and measure the contribution of any quantum stage in the context of the complete computation.
The available project description confirms the experimental implementation and comparisons; it does not provide a public, fully reproducible data package from which to claim a numerical speedup or an operational deployment. The important achievement is the implemented and tested research program, not a performance figure that has not been established publicly.
Continue exploring
Related news: SparQ project announcement