Researchers develop a new method to guide quantum systems before they decay

In everyday life, watching something usually does not change it. Looking at a door does not make it open or closed. Watching traffic does not alter the flow of cars. At the microscopic scale of quantum physics, observation works very differently: measuring a system unavoidably affects its behavior. 

This aspect of quantum physics sits at the heart of new research by scientists at the University of Luxembourg. Prof. Aurelia Chenu and her team, in collaboration with the experimental team of Prof. Kater Murch at WUSL (Washington university in Saint Louis, USA) have developed a way to reliably control quantum systems even while they are being continuously observed, without having to slow the process dramatically. Their work could help make future quantum technologies faster, more reliable, and easier to control.

When observation becomes part of the problem

Many modern quantum technologies rely on continuous monitoring. In superconducting quantum circuits, for example, scientists constantly read out information from the system. Quantum sensors extract signals by observing tiny changes over time. In all these cases, the system is never truly isolated. 

But there is a catch. In quantum mechanics, observation does not just reveal what the system is doing; it also actively influences its evolution. Continuous monitoring can cause sudden and unpredictable jumps in a system’s state, making control difficult. 

To deal with this, physicists often focus only on rare situations where no sudden jump occurs. Along these selected paths, the system can display unusual and highly structured behaviours that do not occur in everyday systems. 

“These trajectories are interesting, but they are fragile,” explains doctoral candidate Niklas Hörnedal. “They rely on delicate measurements and on continuously discarding information when the system behaves unexpectedly.”

Why going slowly usually helps and why it does not here

In many areas of physics, there is a simple rule: if you want to control a system reliably, you need to change things slowly. This gives the system the time to adapt, making its behaviour easier to predict. 

This idea works well for systems that are not being monitored. Under continuous observation, however, slowing down introduces a new problem. The longer the process takes, the more opportunities there are for something to go wrong. Thus, slowing down does not improve control and it makes the desired behaviour extremely unlikely. 

This creates a dilemma: move fast and lose control, or move slowly and almost never succeed. 

A different strategy, counterdiabatic driving

The key breakthrough of the University of Luxembourg researchers was to abandon the idea that reliable control requires slowness. Instead, they introduced an additional layer of control known as counterdiabatic driving, a method that adds a designed correction to the system’s dynamics, so it stays on the intended path even when conditions change quickly.  

Think about driving on a slippery road. One option is to drive extremely slowly and hope nothing goes wrong. Another is to use stability control systems that counter skids in real time, allowing you to move faster while staying in control. Counterdiabatic driving plays a similar role for quantum systems. By designing this extra control, the researchers were able to cancel the unwanted effects that normally appear when system parameters change quickly. As a result, the system follows the desired trajectory at finite speed, spending much less time in the fragile, monitored regime.

‟ Instead of avoiding errors by going slowly or hoping that the system remains naturally stable, we prevent them directly.”

Prof. Aurelia Chenu

Associate Professor in Theoretical physics

View profile

Why this matters

Continuously monitored systems already appear in real technologies today. Superconducting quantum circuits, photonic platforms, and quantum sensors all rely on ongoing measurement. In these systems, the measurement record often contains valuable information, but onlyif the system can be controlled reliably. 

This could open new possibilities. If a system is controlled very precisely, even a tiny change in its environment can have a noticeable effect. This is especially useful for sensing, where such systems could be used to detect extremely weak signals that are otherwise hard to measure. 

In quantum computing and communication, the implications are still unfolding. While most current approaches rely on isolated systems and abrupt measurements, there is growing interest in schemes that involve continuous monitoring, feedback, or selective post processing. Learning how to control these systems could help researchers design new quantum devices that work reliably under real conditions. 

Although the research was demonstrated using quantum circuits, the same control challenge appears in many other settings. It arises in classical systems with gain and loss of energy, such as optical devices, mechanical resonators, or electrical circuits with amplification and dissipation. In these systems, the same control strategy applies, following the same underlying mathematical framework. This means the approach developed at the University of Luxembourg could influence a broad range of physical systems, well beyond quantum mechanics. 

The researchers are already looking ahead. One goal is to study subtle effects that appear when quantum systems are guided along special paths, which can now be observed reliably. They also plan to explore more complex systems, which could behave in entirely new ways when monitored and actively controlled. 

In the longer term, the team hopes their work will help clarify how to guide systems through especially fragile regions, situations where traditional “slow and careful” approaches are known to fail.