SBIR-STTR Award

Aqwest LLC proposes to develop an automated design process using using neural networks and machine learning algorithms, which is adapted for a high-peak-power for a blue laser system with high-repetition-rate. The neural networks-machine learning process will generate blue laser system designs meeting the Navy size, weight, performance, and reliability requirements 50 times faster compared to the conventional manual laser design process. The neural networks-machine learning offers key benefits to the US Navy, namely an automated blue laser design algorithms generating results within 5% variation of the target, which addresses all design parameters of a multistage laser architecture. The algorithms will be verified on a physical prototype and benchmarked to verify time reduction. The resulting algorithms support a range of blue laser architectures including solid-state laser and optical parametric oscillators. They are also adaptable to designing and optimizing of a broad range of laser systems, and extendable to address multi-parameter problems in other fields.

Benefit:
Neural networks and machine learning are emerging groups of technologies offering a low-cost, expedient shortcut to the traditional design / iterative optimization processes at a fraction of time and cost. These technologies will be applied to the blue lasers. The technical advances can be also applied to law enforcement, marine navigation, medical applications, and laser material processing. These conditions create a significant market opportunity.

Keywords:
Neural networks, Neural networks, Machine Learning, laser modeling, Blue laser, laser physics