Two qubits, quantum bits, two-level systems, glide through hot and cold thermal baths. One moves at 0.8 times the speed of light. At that moment, familiar thermodynamics cracks: a moving qubit senses temperature differently than a stationary one. For it, the cold bath can feel hotter, and the hot bath colder. This is not a trick but a direct consequence of relativity, and it turns motion into fuel for a quantum motor.

A classical heat engine is locked in the cage of the Carnot limit, the maximum efficiency that depends only on reservoir temperatures. In the quantum world, we have already learned to break out of that cage by feeding the engine squeezed thermal states. The new work goes further: instead of exotic reservoirs, it uses the motion of the working medium itself. The qubits, acting as Unruh–DeWitt detectors, point-like quantum probes sensitive to the field, race at constant speed, and that is enough. No acceleration, no curved geometry.

The effective temperature, the one felt by the moving object, is computed from the detailed balance of transitions between qubit levels. It depends on the qubit frequency and its speed. In the high-temperature regime (βω ≪ 1), a moving qubit always feels colder than the bath, while in the low-temperature regime (βω ≫ 1), it feels hotter. This opens a paradoxical possibility: an engine that would work as a refrigerator at rest starts producing useful work when moving. The efficiency at maximum power can exceed not only the Curzon–Ahlborn efficiency but also the Carnot limit itself, defined by the rest temperatures. For example, when qubit B moves through the cold reservoir at 0.8c and the temperature ratio is $β_A$/$β_B$ = 1/2, the efficiency reaches values above $η_C$. When the reservoir temperatures are equal, the standard Carnot limit is zero, but the relativistic engine still outputs positive work.

Entropy production remains non-negative, so the second law is not violated. Motion does not create energy from nothing; it redistributes it, much like wind fills sails. Before this work, it was unclear whether inertial motion alone could serve as a resource for improving efficiency without violating the second law. Now we have the answer: it can. Moreover, simple uniform motion suffices, without accelerations or gravitational fields. This simplifies experimental implementation and opens a new class of quantum heat engines where speed control tunes efficiency without changing reservoir temperatures.

Future directions include accounting for the energy cost of maintaining qubit motion and the backreaction of detectors on the field. It would be interesting to generalize to more realistic trajectories, including acceleration, and to interaction with the electromagnetic field instead of a scalar one. One could also study many-particle working media and entangled states to further boost performance. Predictions can be tested in experiments with moving atoms or superconducting qubits that mimic relativistic motion through modulation of the reservoir coupling. The work directly relates to the fundamental problem of defining temperature in relativistic thermodynamics and to the question of efficiency limits for quantum heat engines with non-equilibrium reservoirs.