Skip to main content

Artificial Intelligence/Machine Learning for Photonics, Power & Energy, Atmospherics, and Quantum Science

An Artificial Intelligence (AI) system can learn on its own and perform multiple tasks and functions if trained properly. Machine Learning (ML) is the training of AI systems using statistics and algorithms to enable it to identify patterns in observed data, build models to represent the patterns, and predict things without having explicit preprogrammed rules and models. Deep learning is a subset of ML that attempts to learn in multiple levels, corresponding to different levels of abstraction by devising complex models and algorithms that lend themselves to prediction. Deep learning is synonymous with large and complex neural networks. Optimal control is another area that has benefited from the advancements in deep learning algorithms. With new advances in graphics processing unit technology and new ML algorithms, deep learning methods have been applied to dramatically improve the state of the art in speech recognition, visual object recognition, object detection, and many other domains such as drug discovery, genomics, and robotics. Taking advantage of the inherent fault tolerant nature of neural networks, their ability to extract patterns from large complex datasets, and the ability to converge quickly to an optimal solution, ML techniques can be applied to photonics, power and energy, thermal management and controls, atmospherics, communication, and quantum science for improved naval capabilities.

Research Concentration Areas

  • Directed energy, power and energy, thermal management, controls, atmospherics, and quantum materials
  • AI and ML applied to imaging data to train deep neural networks to improve target image identification and acquisition
  • Deep learning control algorithms applied to large atmospheric datasets to train neural networks to enable atmospheric distortion compensation and real-time optimization of adaptive optics for maximization of laser beam propagation
  • ML algorithms using available power and energy and thermal monitoring datasets for automatic regulation and switching of high power loads on naval vessels and installations for improved efficiency and availability
  • Use of multilayer neural networks to enable quantum computational capabilities and design and operation of sensitive bosonic states in novel materials

Research Challenges and Opportunities

  • Investigate ML techniques for real-time atmospheric distortion compensation of laser beam propagation for adaptive optics for directed-energy systems using data from available sensors
  • Investigate and develop algorithms with data optimization and regularization schemes that work with large datasets for deep learning optimal control problems
  • Conduct theoretical analysis and investigation of deep learning algorithms suitable for automatic control of naval power, energy, and thermal management systems
  • Investigate software/hardware implementation of deep learning algorithms
  • Develop data-driven models for certification of boson sampling devices by using unsupervised methods, such as clustering, to find patterns in high-dimensional data that allow algorithms to learn facts about complex quantum systems. This will enable sensitive quantum detectors and sensors.

How to Submit

For detailed application and submission information for this research topic, please refer to our broad agency announcement (BAA) No. N00014-22-S-B001.

Contracts: All white papers and full proposals for contracts must be submitted through FedConnect; instructions are included in the BAA.

Grants: All white papers for grants must be submitted through FedConnect, and full proposals for grants must be submitted through; instructions are included in the BAA.


Chappell, Sarwat
Program Officer
Code 333