Physicists have designed a set of quantum control protocols that can make a monitored quantum process look as though time is flowing backward, forward at a different speed, or in no clear direction at all. In a new paper (posted on arXiv and highlighted by institutional summaries), the researchers construct a Hamiltonian and feedback scheme that can reshape the measurement-induced arrow of time at the quantum scale. The result challenges the intuition that the direction picked out by a measurement record is always locked in, showing instead that it can be tuned within the trajectory statistics.
Why Measurements Create a One-Way Clock
The microscopic laws of physics are largely symmetric under time reversal. Run a video of two particles scattering off each other, and the reversed clip obeys the same equations. Yet everyday experience, and most quantum experiments, follow a clear forward direction. The standard explanation points to thermodynamics and entropy, but a parallel mechanism operates at the quantum level: the act of measurement itself stamps a direction onto a system’s evolution.
When an observer performs a sequence of measurements on a quantum system, the outcomes form a trajectory, a string of results ordered in time. The probability of seeing that exact string is generally different from the probability of seeing its time-reversed counterpart. That asymmetry is what researchers call the measurement-induced arrow of time, and it has been quantified experimentally using ultracold atoms under continuous observation. The key insight is that the arrow is not merely philosophical; it can be measured through trajectory likelihoods and fluctuation relations borrowed from statistical mechanics.
Separate theoretical work has formalized what a time-reversed protocol actually means in quantum information terms. One approach uses reversed Kraus maps and backward trajectory initialization to define forward versus reversed processes precisely. Without that mathematical scaffolding, claims about “reversing time” would lack operational meaning. The new Physical Review X paper builds directly on this foundation, translating abstract definitions into a concrete control recipe.
How the New Controls Work
The core technique introduced in the new paper is the explicit construction of a Hamiltonian that reproduces the stochastic quantum trajectories produced by monitoring. In plain terms, the researchers figured out how to write down the energy operator that would make a quantum system evolve as if it were being continuously watched, even when the controls are applied through feedback rather than passive observation. That Hamiltonian, combined with a feedback loop conditioned on measurement outcomes, lets an experimenter steer the system’s trajectory so that it becomes more compatible with backward flow of time than with ordinary forward evolution.
The protocol does not simply hit “rewind.” It offers a spectrum of effects. By tuning the feedback, the arrow of time can be stretched, making forward-time behavior more pronounced; blurred so that an observer cannot easily distinguish forward from backward; or nearly inverted so the statistics of the trajectory resemble what would be expected under a time-reversed description. As the authors describe it, the direction is defined operationally through the observer’s measurement record and the associated control protocol (in their framework at the Joint Center for Quantum Information and Computer Science at the University of Maryland and NIST).
This matters because classical randomness is understood to come from incomplete knowledge of a system’s microscopic state. Quantum randomness is different: it is intrinsic, arising from measurement itself. The new controls exploit that distinction. Because the arrow of time in quantum systems is generated by the measurement process, intervening in that process with the right Hamiltonian and feedback can reshape the arrow rather than merely observe it. In effect, the measurement “clock” becomes another knob on the control panel.
Earlier Attempts to Reverse Time’s Direction
The idea of running quantum processes backward has attracted attention before, but earlier demonstrations operated under tighter constraints. A well-known experiment implemented a time-reversal procedure on IBM’s public quantum computer, described in a paper on few-qubit dynamics. That work generated wide coverage that portrayed it as reversing the flow of time. The reality was narrower. As Argonne National Laboratory clarified in its own account, the reversal involved carefully engineered evolution in a small Hilbert space, not a macroscopic rewind. The algorithm worked on a handful of qubits and required precise knowledge of the initial state.
Other researchers have explored universal time-reversal protocols that aim to rewind unknown unitary operations. That line of work distinguishes between probabilistic reversal, which succeeds only some fraction of the time, and protocols that approach deterministic reversal for two-level systems. The distinction is important: a protocol that works only probabilistically has limited practical use for error correction or state preparation, where reliability is crucial. These schemes also typically assume that the system’s evolution is unitary and isolated from measurement backaction.
A separate strand of research has shown that time’s arrow can be placed in a quantum superposition. In certain thermodynamic setups, forward and time-reversed protocols can be combined coherently, with decoherence or projective measurement selecting an apparent direction. That finding suggested the arrow of time is not as rigid as classical intuition implies, but it stopped short of offering a general-purpose control tool. The superposition experiments were demonstrations of principle rather than blueprints for engineering devices.
What Changes With Controllable Time Direction
The new work goes further than any single predecessor by combining Hamiltonian construction with feedback into a unified control framework. Rather than reversing a known unitary or superposing two fixed protocols, it treats the arrow of time as a tunable parameter. That shift has concrete implications for quantum technology.
One potential application is energy extraction. When a quantum system is monitored, the measurement process can inject or remove energy depending on the trajectory. A Los Alamos National Laboratory release tied to the research notes that, in principle, using the constructed Hamiltonian in a feedback loop could allow an experimenter to extract work from the monitoring process itself, effectively turning information in the measurement record into a thermodynamic resource. In this picture, reshaping time’s direction is not just a conceptual trick; it can change the predicted work/energy balance associated with monitored quantum dynamics.
Another potential use lies in quantum error correction and metrology. Many error-correcting codes and sensing schemes already rely on continuous monitoring and feedback. If the same hardware can be repurposed to bias trajectories toward or away from their time-reversed counterparts, engineers may be able to suppress certain classes of noise more efficiently. For example, protocols that make forward-time statistics especially distinctive could help distinguish genuine signals from spurious fluctuations that mimic reversed trajectories.
The framework also clarifies foundational questions. For decades, debates about the arrow of time have pitted thermodynamic explanations against quantum-mechanical ones. By showing that the measurement-induced arrow can be dialed up, down, or flipped through explicit controls, the new work suggests that time’s direction is, at least in part, an emergent and manipulable property of the measurement record. The underlying equations remain symmetric; what changes is how observers choose to intervene.
Limits and Open Questions
Despite its conceptual reach, the protocol has limits. It does not allow anyone to send information into the past or undo macroscopic events. The apparent reversal occurs within the statistics of quantum trajectories, not in the everyday world of broken eggs and aging stars. Implementing the required Hamiltonian and feedback in a laboratory will also be technically demanding, especially for systems larger than a few qubits or continuous variables with many accessible states.
There are open questions about robustness. Real experiments involve decoherence, imperfect detectors, and delays in feedback loops. How far the arrow of time can be reshaped under such non-ideal conditions remains to be seen. It is also unclear how the protocol scales: does controlling the arrow in one part of a multipartite system constrain what can be done elsewhere, or can different subsystems be tuned independently to point in different temporal directions?
Even with these caveats, the work marks a shift in perspective. Instead of treating the arrow of time as a backdrop against which quantum events unfold, it promotes that arrow to a controllable degree of freedom. Whether future devices use this capability for energy harvesting, error suppression, or new forms of quantum simulation, the message is the same: at the smallest scales, time’s direction is less a given and more a dial that, with sufficient ingenuity, can be turned.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
