SBIR-STTR Award

To address CBDs need for development of analytical approaches to fuse sensor data from multiple modalities to provide common operational pictures in threat environments, Physical Optics Corporation (POC) proposes to develop a Wavelet-based Approach for Signal Processing and Analysis (WASP). WASP is based on novel application of wavelet-based analysis to multimodel, multidimensional data available from a wide variety of threat sensors. WASP will enable the determination of key features of the data at various levels of resolution and provide the ability to move back and forth between higher and lower resolutions, as well as dimensions, as needed. This multiresolution approach enables multilevel decision making and incorporation of sensors providing data at different resolutions. The capability to provide data features using multiple signature feeds and threat information with various levels of detail, which directly addresses the CBDs need for a comprehensive situational understanding regarding the presence of threats in an operational environment. In Phase I, POC will demonstrate the feasibility of WASP by implementing a TRL-4 prototype demonstration using the data generated by POC sensors. In Phase II, POC plans to further develop WASP to TRL-5/-6 by optimizing algorithms for efficiency and accuracy for eventual integration with a program of record.

To address the CBDs need for a signal processing system designed to analyze and fuse data collected from chemical and biological sensors, Physical Optics Corporation (POC) proposes to further develop the Wavelet-Based Approach for Signal Processing and Analysis (WASP). WASP is based on novel application of Wavelet-based analysis to multimodal, multidimensional data available from a wide variety of threat and meterological sensors. WASP algorithms filter threat data to isolate key features, perform multiresolution analysis to eliminate noise and detect changes to obtain probabilistic models of the threat environment. The probabilistic models are used to fuse data, yielding a common operational picture of the threat environment and sensor retasking recommendations. WASP thus enables multimodal data-fusion, and multilevel decision making and provides the ability to compute source-terms from chemical/biological and meteorological sensors available at different spatiotemporal resolutions, thus directly addressing the CBDs need for a comprehensive situational understanding regarding the presence of threats in an operational environment. Phase II work will mature WASP algorithms successfully developed during Phase I, incorporate a standard modular Common CBRN Sensor Interface (CCSI)-compliant framework, and customize WASP for eventual integration with a program of record such as JWARN, JOEF, and JEM, yielding a TRL-6 system. ---------- To respond to the U.S. military requirement for new tools to provide early warning on the existence of chemical/biological contamination in the battlefield, Physical Optics Corporation (POC) proposes to further develop Plume Analysis and Classification for Early Warning (PEACE). PEACE is designed to process LIDAR detection of plume clouds and extract key information by determining the plume’s content and estimating the method used to release chemical/biological contaminants in the environment. PEACE uses machine learning (ML) and artificial intelligence (AI) algorithms to detect contaminant material type and release method. ML/AI algorithms classify the plumes into a set of known contaminants and compute the confidence levels for the classification results. PEACE improves the classification accuracy by selecting ML/AI algorithm parameters (such as number of distinguishable classes of contaminants) that are sensitive to ambient conditions and fine-tuned to contaminant(s) of interest. Under this Sequential Phase II project, the PEACE algorithms successfully developed during Phase II project WASP will be matured further by refining the ML/AI algorithms, incorporating a standard modular Common CBRN Sensor Interface (CCSI)-compliant framework, and demonstrating performance improvements by using an extensive set of data. PEACE is designed for eventual integration with a Stryker sensor suite.