CDI’s cross-disciplinary team of PNNL researchers is developing the science to address inverse problems in chemical systems. This challenge, while on its own difficult, is made more so by the sparse data environments that often accompany the measurement of chemical signatures in many national security mission areas.

The CDI team is investigating how new tools and techniques in measurement, uncertainty quantification, data science, and chemical theory can lead to a better understanding of the fate of real-world chemical signatures. One of the primary questions is, “What is the data density required to leverage new tools such as artificial intelligence and machine language?”

The CDI team will answer this question and others by developing an infrastructure and mechanisms to embed fundamental chemical and physical principles into cutting-edge data science tools. This requires the application of first-principles theory to develop synthetic and simulated data to fill in gaps within sparse data sets, as well as cutting-edge measurement science to generate high information content data.

CDI research activities fall under three use cases, as well as providing integration through the Data and Integration Activities:

Nuclear Incident Characterization: Small molecule indicators of nuclear accidents that are subject to changes in speciation, volatility, and fractionation when released into the environment.

Additive Manufacturing: Complex chemical filaments that go through a highly engineered manufacturing process to form new structures and devices using additive manufacturing.

Functional Materials: Materials and devices that have been constructed with atomic precision from chemical precursors to operate as sensors in extreme environments.

Data and Integration Activities: Develop a common framework across projects to study chemical signature dynamics; improve the integration of models and measurements within a coherent mathematical framework.