APPLIED RESEARCH LLC — Department of Energy STTR Phase II: C52-32e

Description

Radiation compliance is generally demonstrated by monitoring radiation fields both onsite and offsite with both passive (dosimetry) and active, real-time monitors. Conventional approachesfor demonstrating compliance rely on the expertise of individuals for interpreting data, configuring proper estimations, performing calculations, etc. Additionally, changes in facility configurations requires significant personnel effort in configuring equipment, reanalyzing data, and rerunning calculations. A low cost, automatic, and accurate approach is to create a software package that contains data acquisition, data preprocessing, machine learning (ML) and artificial intelligence (AI) algorithms for interpreting the data, and a user friendly graphical user interface (GUI). In Phase I, we used actual data obtained from JLab to validate our algorithms. The first step is to apply advanced deep learning and machine learning algorithms to predict radiation levels. Several conventional and deep learning algorithms were developed. Our results clearly showed that the radiation prediction is very accurate. The benefit of these results is significant; the results imply that in cases where a detector fails to respond, or whose response is erratic, the estimated measurement result can be reconstructed using a simple model similar to what was implemented. The second step is to construct a dense radiation map based on very sparse sensor measurements in a given laboratory. More than five conventional and one deep learning based inpainting algorithms have been investigated. The dense maps generated by using one of the inpainting algorithms are also accurate. Therefore, we have successfully demonstrated the feasibility of the proposed approach. In Phase II, we will develop a software prototype containing radiation prediction algorithms, dense radiation map algorithms, and background radiation prediction algorithms. Actual data will be used to evaluate the prototype. Our direct customers include the DOE facilities, nuclear power plants, medical research facilities, and possibly nuclear submarines. There are also several potential markets for our ML technologies. The global Internet of Things (IoT) market was valued at $5,533.25 million in 2020 and is estimated to grow to $18,046.71 million in 2025, at a CAGR of 26.67% from 2020 to 2025. The radiation detection industry, which was valued at USD 2.85 billion in 2020, and is expected to reach USD 4.13 billion by 2026, registering a CAGR of 6.67% during the forecast period of 2021-2026. The radiation dose monitoring market would grow from $77.01 million in 2020 to $139.7 million in 2025, at a CAGR of 12.65% from 2020 to 2025. The global AI software market revenue grows at a CAGR of 41.6%, from $14.7 billion in 2019 to $118.6 billion in 2025. Given the strong expertise of our team, the likelihood for success commercialization to the above markets is extremely high.