The challenge of invisible errors

Quantum computers are inherently sensitive systems. Interactions with the surrounding environment can introduce errors that disrupt calculations long before they finish.

Most of these errors occur ‘silently’. A qubit may lose energy or drift away from its intended quantum state without any indication that something has gone wrong. Because the error location is unknown, complex error-correction techniques are required to identify and repair the damage.

These techniques are powerful, but they come with significant overhead in both hardware and computational resources. For quantum computing to scale efficiently, reducing this overhead is essential.

An alternative approach is to design qubits so that the most common type of error is a detectable, or ‘heralded’, event known as an erasure. In an erasure error, the system doesn’t just subtly drift from its intended state, but leaves the computational space entirely in a way that can be explicitly observed. When this happens, the system effectively raises a flag saying that this result cannot be trusted.

This information is extremely valuable. If an erasure can be reliably detected, the system can simply discard that run of the computation and repeat it, rather than trying to infer and correct the mistake afterwards. This technique is called postselection.

For certain types of quantum hardware, including architectures based on dual-rail qubit encodings, this type of error detection can emerge naturally from the physics of the system.

Turning errors into detectable events

One way to reduce the burden of error correction is to change the way quantum information is encoded.

In our work, we explore qubits encoded across two physical modes, sometimes referred to as dual-rail qubits. Instead of storing information in a single component, the quantum state is distributed across two.

This structure introduces an important advantage. If one of the physical elements loses energy, one of the most common types of error in superconducting quantum hardware, the system can transition into a state outside the valid computational space. Crucially, this transition can be detected via a measurement which doesn’t otherwise affect the logical state.

When this happens, the system effectively raises a flag indicating that the computation has been compromised and , rather than allowing the error to propagate through the rest of the circuit, the affected run can be discarded and repeated.

Measuring the benefits

Detecting errors in this way can significantly improve the reliability of quantum computations. However, it also introduces a trade-off. Every discarded run means the computation must be repeated. If errors occur too frequently, repeatedly restarting the circuit can quickly become costly.

To understand when this approach is beneficial, the OQC team has developed simulation tools to model erasure noise and postselection in quantum circuits. These tools allow researchers and customers to explore how error rates, circuit complexity, and detection performance influence the overall efficiency of a quantum computation.

As part of this work, we have released an open-source simulation library, erado, built on top of Qiskit. This enables researchers to model erasure-aware noise and investigate how postselection strategies impact different quantum algorithms.
The results show that erasure detection could improve the reliability of quantum computations under certain conditions: particularly when the dominant errors are detectable and when detection itself is highly reliable. However, the benefits are not unlimited. If detection is imperfect or if the rate of erasure events becomes too high, the cost of repeated postselection can grow rapidly.

Understanding these limits is essential for guiding the design of next-generation quantum hardware and algorithms. In other words, the value of erasure-aware architectures lies not only in the concept itself, but in knowing when and where they provide the greatest advantage.

Designing the next generation of quantum systems

The results of this study highlight both the promise and the limitations of erasure-based approaches. When error detection is reliable and the rate of erasure events remains manageable, postselection can significantly improve the quality of quantum results. But the benefits depend strongly on hardware design and system performance.

Understanding these trade-offs is critical for guiding the development of future quantum architectures. Research like this helps inform how quantum hardware, qubit encodings, and error-mitigation techniques can work together to deliver more reliable computation.

Advancing practical quantum computing

At OQC, our mission is to build scalable quantum computing systems that deliver real-world value. This requires tackling the practical challenges of operating quantum processors, including how errors arise, how they propagate through circuits, and how they can be detected or mitigated.
By exploring new approaches such as erasure-aware qubit encodings and postselection, at OQC we are developing a deeper understanding of how quantum systems can be made more reliable and more useful. As quantum hardware continues to advance, insights like these will help shape the architectures and techniques that underpin the next generation of quantum computers.

For a conceptual introduction, usage instructions and API reference, and more information:

REPO: https://github.com/oqc-community/erado
DOCS: https://oqc-community.github.io/erado
PYPI: https://pypi.org/project/erado