Continuously Monitored Quantum Sensors: Smart Tools and Applications

获取和解释有关物理过程的数据对科学和技术至关重要。“连续监测量子传感器：智能工具和应用”（C'MON-QSENS!）项目的目标突破是开发用于解释从量子传感器获取的数据的工具。事实上，量子增强的超精密传感器是最具颠覆性的量子技术之一，在几个学科中有短期应用，但到目前为止还很有限。大多数工作都致力于通过单次或重复测量方案测量静态特性，而许多实际应用都与动态信号有关。从时间序列数据中提取信息需要传感器在持续监测状态下运行，这就是C'MON-QSENS的跨学科方法！出现了。我们的目标是开发连续监测的热原子系综和光力学器件，并与量子信息论、统计推断和经典信号处理领域的领先实验家和理论研究人员合作，追求它们的应用。我们将创造一种独特的协同作用，以缩小跨学科的差距，因此（经典）信号处理和数据推断的现代方法可以在量子计量的背景下被纳入。该结果将允许在实验中明确演示先进的传感任务。我们将对实时状态下的量子信息处理有更深入的了解，并开发量子传感和实时信号解释的实用方法。C 'MON-QSENS !A.构建先进的动力学模型，以便准确描述实时量子传感器，包括相关的退相干机制、非线性、随机噪声源和由连续时间测量产生的量子反作用。B.开发(i)信号处理和统计推断技术（贝叶斯滤波，压缩感知，序列分析），用于高度控制的场景，当量子传感器和信号动态可以精确建模时，以及（ii）用于现实世界复杂场景的无模型机器学习方法。这将推进连续监测量子系统的基础理论，并为相关传感任务的性能提供最终界限。基于连续监测原子蒸汽和光力学系统构建量子传感器。我们将应用动态模型和推理技术来优化传感器的运行机制，以允许跟踪现实生活中的信号（例如神经元，大脑，心脏和加速度），并验证先进的传感任务，如波形估计，模型选择和变化点/异常检测。

Acquiring and interpreting data about physical processes is vital for science and technology. The targeted breakthrough of the project "Continuously Monitored Quantum Sensors: Smart Tools and Applications" (C'MON-QSENS!) is to develop tools to interpret data acquired from quantum sensors. Indeed, quantum-enhanced ultra-precise sensors are among the most disruptive quantum technologies with near-term applications in several disciplines, but with a limited reach so far. Most efforts are devoted to the measurement of static properties by single-shot or repeated measurement schemes, while many real-world applications are concerned with dynamical signals. Extracting information from time-series of data needs sensors operating in the continuously monitored regime, and here is where the interdisciplinary approach of C'MON-QSENS! emerges. We aim to develop continuously monitored hot atomic ensembles and optomechanical devices, and we pursue their application in a collaboration with leading experimentalists and theory researchers in quantum information theory, statistical inference and classical signal processing. We will create a unique synergy to close the interdisciplinary gap, so modern methods of (classical) signal processing and data inference can be incorporated within the context of quantum metrology. The result will allow advanced sensing tasks to be explicitly demonstrated in experiments. We will both gain a deeper understanding of quantum information processing in the real-time regime, and develop practical approaches to quantum sensing and interpretation of real-time signals. C'MON-QSENS! will advance the current frontiers of fundamental and applied knowledge on continuously monitored quantum systems by: A. Constructing advanced dynamical models to allow for an accurate description of real-time quantum sensors, including relevant decoherence mechanisms, non-linearities, sources of stochastic noise, and quantum back-action resulting from continuous-time measurements. B. Developing (i) signal processing and statistical inference techniques (Bayesian filtering, compressed sensing, sequential analysis) for highly controlled scenarios when the quantum sensor and signal dynamics can be accurately modelled, and (ii) model-free machine learning methods for real-world complex scenarios. This will advance fundamental theory on continuously monitored quantum systems and provide ultimate bounds on the performance for the relevant sensing tasks.C. Building quantum sensors based on continuously monitored atomic vapours and optomechanical systems. We will apply the dynamical models and inference techniques to optimize the sensors' operating regimes to allow tracking of real-life signals (e.g. neuron, brain, heart, and acceleration) and validate advanced sensing tasks such as wave-form estimation, model selection and change-point/anomaly detection.