Publications and Resources

We designed, implemented, and tested a solution that uses a robotic arm control system capable of picking soil in an outdoor environment.

We designed, developed, and tested a prototype system to automate a bench-top centrifuge, which has applications in reducing the cost and number of man hours for lab sample processing.

Noblis' EMBER is a wildfire activity predictive model tailored to end-user needs for decision making on timescales longer than one week.

Noblis' good and grounded (G3) fact verification solution is an emerging approach to counter the threat of misinformation generated by large language models by malicious or unwitting actors.

Our OptiSource solution uses advanced modeling and simulation to forecast workloads and resources for surface ship drydocking and other maintenance periods.

In support of the Navy’s logistics, this research project evaluated automation opportunities to increase operational efficiency and resiliency of readiness in a contested logistics environment.

Our novel deep learning approach enables automation and scalability to identify new malware that is outside known threat actors.

Our architecture uses machine learning (ML) and enables improved accuracy for long-term horizon weather forecasting and enhanced operational planning capabilities.

Our approach links real-time observations, geospatial analytics, and predictive modeling to improve decision-making workflows.

A phased program involving advanced site characterization approaches and tools is required to complete remedial investigations of complex PFAS sites.

Investigating the atmospheric and surface characteristics driving one of California’s largest wildfires towards development of AI/ML-based early warning systems.

Autonomous sampling machines need the ability to target samples despite limited visibility and robotic arm reach distance. We design a method to speed up the search process.