RII Track-2 FEC: Harnessing the Data Revolution for the Quantum Leap: From Quantum Control to Quantum Materials

量子信息科学有望在量子计算、网络、隐私和传感领域提供变革性的应用。除了量子信息科学对于理解基础量子科学的重要性之外，量子信息的进步对于未来信息社会的国家安全和经济都具有战略意义。然而，随着越来越大、越来越复杂的量子设备的构建，一个关键的挑战是以一种保持其脆弱的量子性质的方式来控制它们。为了达到所需的控制水平，必须精确识别量子系统、材料或过程的关键属性和特征。该项目通过使用自举方法来解决关键属性的识别问题，将当今的小型量子计算机与大规模经典计算资源相结合，以设计下一代量子计算机。这种方法将允许系统地改进大型量子系统，并设计具有更好功能特性的设备，例如固有的抗错误能力。具体来说，该项目将结合使用机器学习来筛选具有所需特性的候选材料，使用现有的中等规模量子处理器对有前途的系统进行量子模拟，以改进自适应学习策略，并对基本微观材料特性进行实验验证。这项研究的变革性目标是开发改进的大规模量子系统的鲁棒和精确控制。通过整合大数据、量子模拟和实验验证来解决量子信息科学中的基本挑战，该项目旨在协同利用这些多样而强大的工具所带来的好处。开发的合作和技术将建立一个独特的量子信息科学卓越中心，以响应公认的国家优先事项。当地的基础设施与训练有素的量子知识劳动力相结合，将有助于确保美国在量子技术发展方面的竞争力。EPSCoR提案汇集了来自布朗大学（RI）和达特茅斯（NH）的研究人员团队，以研究使用新的数据科学方法来解决量子科学中的两个关键挑战：（i）复杂系统的系统识别和量子控制;量子材料的多体模拟。随着越来越大的量子系统的构建，一个关键的挑战是精确地识别系统的哈密顿量（更一般地说，是底层的动力学模型），并根据需要精确地操纵它。拟议工作的一个关键特征是使用量子自举来系统地完善我们对量子多体系统的理解，并设计具有所需功能特性的新系统，例如可能允许量子信息编码的拓扑保护状态。这项工作涉及机器学习的组合，以筛选具有所需特性的候选材料，使用中等规模量子处理器对有前途的系统进行量子模拟，以改进自适应学习策略，以及对基本微观材料特性进行实验验证。合作研究的变革性目标是通过使用高度可控的磁共振技术获得的实验数据进行算法学习，开发哈密顿和开放系统（例如Liouvillian）识别方法来表征未知的量子系统。该项目将开发工具来解释环境噪声，以实现对量子动力学的鲁棒、高保真控制。合作和开发的技术将允许在美国建立一个独特的量子信息科学研究中心，具有长期研究能力。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

Quantum information science is poised to deliver transformative applications in the areas of quantum computing, networking, privacy, and sensing. In addition to the importance of quantum information science for understanding of basic quantum science, quantum information advances have strategic relevance for both national security and the economy of future information-based societies. However, as ever larger and more complex quantum devices are constructed, a key challenge is to control them in a way that preserves their fragile quantum nature. To achieve the required level of control, it is essential to precisely identify crucial properties and features of the quantum system, material, or process of interest. This project addresses the identification of key properties by using a bootstrapping approach, combining today's small quantum computers with large-scale classical computing resources to design the next generation of quantum computers. This approach will allow to systematic refining of large quantum systems, and engineering of devices with better functional properties such as intrinsic resistance to errors. Specifically, the project will use a combination of machine learning to screen candidate materials with desired properties, quantum simulation of promising systems using available intermediate-scale quantum processors to refine adaptive learning strategies, and experimental validation of the fundamental microscopic material properties. The transformative goal of this research is to develop improved robust and accurate control of large-scale quantum systems. By integrating big data, quantum simulation, and experimental validation to solve fundamental challenges in quantum information science, the project aims to synergistically leverage the benefits offered by these diverse and powerful tools. The collaborations and techniques developed will build a unique center of excellence for quantum information science, in response to a recognized national priority. The local infrastructure combined with a highly trained quantum-literate workforce will be instrumental in ensuring American competitiveness in quantum technology development.This EPSCoR proposal brings together a team of researchers from Brown University (RI) and Dartmouth (NH) to investigate the use of novel data science methods to address two key challenges in quantum science: (i) System identification and quantum control of complex systems; and (ii) Many-body simulation of quantum materials. As ever larger quantum systems are constructed, a key challenge is to precisely identify the system Hamiltonian (even more generally, the underlying dynamical model) and to precisely manipulate it as desired. A key feature of the proposed work is to use quantum bootstrapping to both systematically refine our understanding of a quantum many-body system, and to engineer novel systems with desired functional properties, such as topologically protected states that may permit encoding of quantum information. This effort involves a combination of machine learning to screen candidate materials with desired properties, quantum simulation of promising systems using intermediate scale quantum processors to refine adaptive learning strategies, and experimental validation of fundamental microscopic materials properties. The transformative goal of the collaborative research is to develop both Hamiltonian and open-system (e.g. Liouvillian) identification approaches to characterize unknown quantum systems, by using algorithmic learning with experimental data obtained by highly controllable magnetic resonance techniques. The project will develop tools to account for environmental noise to enable robust, high-fidelity control of quantum dynamics. The collaborations and techniques developed will allow building of a unique center for quantum information science research in the US, with long-term research capabilities. The participating graduate and undergraduate students will form a valuable quantum-literate workforce.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.