| SCAN | |
|---|---|
| Name | SCAN |
| Caption | Schematic representation of SCAN protocols in quantum measurement |
| Field | Quantum physics; Quantum information science |
| Related | Quantum sensing, Quantum measurement, Quantum tomography, Quantum control |
| Developer | Various research groups in quantum information and quantum optics |
| Introduced | 21st century |
SCAN
SCAN is a family of measurement and analysis techniques in quantum physics emphasizing adaptive, compressed, or structured approaches to extract information from quantum systems with reduced resources. SCAN protocols matter because they aim to reconcile limited experimental access and noise with the demands of high-fidelity state estimation, metrology, and control in contemporary quantum computing and quantum sensing platforms.
SCAN broadly stands for methods that use Structured/Sequential/Selective Compression, Contrastive Adaptive Networks, or similar acronyms depending on subfield conventions; the term has been used to label protocols that combine experimental scanning strategies with algorithmic compression and adaptive decision rules. Typical SCAN workflows integrate prior models (e.g., low-rank structure or sparsity), sequential measurement scheduling, and statistically efficient estimators to reconstruct states or characterize dynamics with fewer measurements than naive full tomography. SCAN is connected to areas such as compressed sensing, adaptive measurement, and machine learning applied to quantum experiments; it is often deployed alongside techniques from quantum tomography and Hamiltonian learning.
Foundations of SCAN draw on quantum estimation theory, information-theoretic limits, and convex optimization. Key theoretical pillars include the Cramér–Rao bound, Fisher information, and resource-counting in quantum metrology as formalized in works by researchers at institutions such as Perimeter Institute, MIT, Caltech, and University of Cambridge. SCAN analyses commonly assume low-rank density matrices or sparse process matrices and exploit results from compressed sensing (notably work by Emmanuel Candès and Terence Tao in classical contexts) and their quantum generalizations (e.g., David Gross et al.). Bayesian and frequentist estimators, including adaptive Bayesian experimental design and sequential Monte Carlo methods, underpin optimal measurement allocation in SCAN. Connections to matrix product states and tensor-network descriptions permit SCAN-style inference for many-body systems when entanglement is structured, linking to research by groups at Max Planck Institute for Quantum Optics and ARC Centre of Excellence in Engineered Quantum Systems.
Experimental SCAN implementations appear across platforms: superconducting qubits (e.g., IBM Quantum, Google Quantum AI testbeds), trapped ions (e.g., NIST, IonQ), cold atoms and optical lattices (e.g., JILA, CERN collaborations in quantum simulation), and solid-state defects (e.g., NV center experiments at University of Basel). Techniques include adaptive measurement settings chosen by real-time controllers, compressed readout bases implemented with randomised measurements (linked to protocols by Philipp H. Häffner and colleagues), and neural-network-based reconstruction (e.g., restricted Boltzmann machines and variational autoencoders) trained on SCAN-generated data. Practical concerns—state preparation and measurement (SPAM) errors, decoherence, and finite sampling—are mitigated through cross-validation with benchmark protocols such as randomized benchmarking and gate set tomography developed by researchers at Sandia National Laboratories and IBM Research.
SCAN methodologies serve multiple roles: efficient quantum state and process tomography for quantum processors, Hamiltonian parameter estimation for quantum simulators, and enhanced sensitivity in quantum metrology. In quantum computing, SCAN reduces characterization overhead for mid-scale devices where full tomography is infeasible, supporting calibration workflows used by Rigetti and cloud quantum services. In sensing, SCAN-inspired compressive readout enhances protocols in quantum magnetometry and quantum imaging using NV centers and cold-atom interferometers, improving equitable access to precision measurement by lowering hardware and data burdens. SCAN also facilitates resource-aware protocols for quantum error mitigation and verification tied to proposals from John Preskill and other theorists advocating pragmatic fidelity estimation in near-term devices.
SCAN is one approach among complementary frameworks: full quantum tomography, direct fidelity estimation (as advanced by Eisert-style proposals), randomized measurement techniques (e.g., classical shadows developed by Huang, Kueng, and Preskill), and dedicated Hamiltonian learning algorithms. Compared with classical shadows, SCAN emphasizes adaptive and structure-exploiting measurement schedules rather than fixed random ensembles; compared with variational tomography it stresses principled sampling strategies rooted in statistical decision theory. Choice among these frameworks depends on target state complexity, noise models studied by groups at Los Alamos National Laboratory and University of Chicago, and practical constraints such as measurement parallelism and classical post-processing resources.
SCAN research intersects with broader questions of equitable access to quantum technologies and responsible allocation of scientific resources. By lowering experimental and computational costs for characterization, SCAN can democratize participation in quantum research across under-resourced institutions and Global South laboratories. However, biases in training datasets, proprietary toolchains from corporate actors (e.g., Google and IBM), and concentration of infrastructure risk reinforcing inequalities; open-source SCAN toolkits and community-driven benchmarking (advocated by consortia like the QED-C and academic coalitions) aim to promote transparency. Ethical considerations also include dual-use potentials of enhanced sensing and surveillance capabilities; policy discussions at national labs and forums such as AAAS emphasize governance, inclusive workforce development, and equitable distribution of benefits from quantum advances.
Category:Quantum physics methods Category:Quantum information science