Superconducting qubits are vulnerable to ionizing radiation, including γ-rays from environmental radioactivity and cosmic-ray muons. These particles generate a large number of phonons in the chip substrate, leading to qubit decoherence and correlated errors across multiple qubits—posing a serious challenge for fault-tolerant quantum computing. While shielding can suppress most radioactive backgrounds, cosmic-ray muons require either modifications to chip design or active mitigation strategies to enable operation in above-ground environments.
We present the development and first experimental validation of a muon tagging system based on a stack of Kinetic Inductance Detectors (KIDs), designed for integration with quantum processors. Operating at millikelvin temperatures, our prototype demonstrates efficient muon tagging (~90%) through coincident signals in a three-layer KID array. We measure a tagging rate of (193 ± 7) × 10-3 events per second, consistent with Monte Carlo expectations, while preserving compatibility with quantum chip packaging and cryogenic constraints.
This system provides a viable path toward active veto architectures capable of identifying and rejecting muon-induced events in real time, and thereby suppressing correlated error bursts. Our results support a hybrid strategy of detection and mitigation, paving the way for scalable and resilient quantum computing platforms.