Capabilities
ARiA applies broad interdisciplinary experience in acoustics, signal processing, machine learning / artificial intelligence, modeling & simulation, and cognitive science toward innovative basic and applied science and engineering research and development.
We specialize in developing:
- real-time physics-based modeling algorithms for simulation and training
- machine-learning algorithms for automated pattern recognition
download a PDF of the ARiA Capabilities Statement
Array, Signal, and Information Processing
ARiA maintains broad expertise in the entire processing chain of antisubmarine warfare (ASW) and mine countermeasures (MCM) sonar systems from processing of the sensor-array data, to signal processing for match filtering, normalization, and detection, to information processing for classification and tracking. We both work to improve the performance of multiple fielded tactical systems and develop novel new approaches including compressive sensing and topological signal processing.
Representative projects:
- Distributed sensor fusion for tactical ASW systems to enhance detection and minimize clutter
- Sparse processing for predicting transmission loss (TL) from small numbers of tactical measurements made by a distributed sensor field used to predict sonar-system performance
- Topological signal processing for classification of MCM sonar echoes
- Sparse processing for tactical ASW systems to enhance detection and minimize clutter
- Compressive and model-based beamforming for tactical ASW systems to enhance detection and minimize clutter
- Adaptive matched-filtering, detection and normalization for tactical ASW systems to enhance detection and minimize clutter
Customers:
- Naval Air Systems Command (NAVAIR)
- Naval Air Warfare Center Aircraft Division (NAWCAD)
- Naval Sea Systems Command (NAVSEA)
- Office of Naval Research (ONR)
ARiA recognizes that “the best pattern recognition system that we know of is our auditory system” and has developed biomimetic signal-processing algorithms for detection and classification and machine-learning algorithms that learn from listeners how to best classify and identify complex signals. We also recognize that physics-based approaches can give unique insight to complex problems of detection and classification. At the same time, we create models that are data-driven and learn from the data themselves in order to avoid errors and brittleness that can arise from model-data mismatches.
Game-Based and Simulation-Based Training
ARiA brings combined expertise in game development and physics-based simulation together in our work. An early adopter of game platforms for simulation-based training, we have integrated research models in commercial game engines and deployed commercial-grade games on military tactical systems.
Representative projects:
- WaveQuest: game-based education and training on fundamental concepts of underwater acoustics
- ESAIL: 3D visualization platform for antisubmarine warfare (ASW) training on tactical systems using simulation, data reconstruction, and “what if…?” analysis
- AIRS: portable virtual-reality simulation and game-based spatial-awareness training for mitigation of spatial-disorientation in pilots and nonpilot crew
Customers:
- Naval Air Systems Command (NAVAIR)
- Naval Air Warfare Center Training Systems Division (NAWCTSD)
- Naval Sea Systems Command (NAVSEA)
- Office of Naval Research (ONR)
Machine Learning and Artificial Intelligence
ARiA maintains broad expertise in machine learning and artificial intelligence for multiple domains – ranging from synthetic-aperture imagery to legal documents – using a variety of supervised, unsupervised, and semi-supervised techniques including statistical-learning approaches ranging from traditional algorithms to cutting-edge deep-learning techniques as well as Bayesian and formal-logic approaches.
Representative projects:
- Deep learning for explainable classification of synthetic-aperture images and parametric synthesis of realistic synthetic-aperture images
- Deep learning algorithms for predicting transmission loss from small numbers of tactical measurements made by a distributed sensor field used to predict sonar-system performance
- Synthetic ASW Generation Engine (SAGE): Combination of statistical machine learning and formal logic reasoning for generation of training scenarios that meet training goals with a constrained computational load
- Machine Interface for Contracting Assistance (MICA): natural-language question-answering system for federal acquisition regulations based on semi-supervised statistical-learning algorithms and formal logic
Customers:
- Air Force Office of Transformational Innovation (SAF/AQ)
- Air Force Research Laboratory Information Directorate (AFRL/RI)
- Naval Air Systems Command (NAVAIR)
- Naval Sea Systems Command (NAVSEA)
- Office of Naval Research (ONR)
Computational Acoustic and Psychoacoustic Modeling
ARiA maintains broad expertise in formulating models of sound propagation and scattering in diverse environments – ranging from underwater, to outdoors, to interior spaces – and human perception of those sounds including detection and localization. Our computational models, developed in both the time and frequency domains, use in-house software implementations that leverage multi-core CPU and many-core GPU architectures to optimize performance.
Representative projects:
- Edge-source modeling of midfrequency target scattering for development of signal-processing algorithms and prediction of sonar performance
- Underwater acoustic modeling in uncertain environments for prediction of sonar performance
- Real-time acoustic simulation of underwater propagation for training applications and tactical decision aids
- Simulation of synthetic targets in real data using physical modeling for prediction of information-processing algorithm performance
- Helicopter noise propagation modeling and modeling of human detection and localization in outdoor environments
- Bullet-noise propagation modeling and modeling of human detection and localization in out door environments
Customers:
- Department of Justice (DoJ)
- Naval Air Systems Command (NAVAIR)
- Naval Sea Systems Command (NAVSEA)
- Office of Naval Research (ONR)
- Supreme Court of the State of New York
Our broad areas of expertise in computational acoustic and psychoacoustic modeling include
- Physical and computational models for propagation, reverberation, and scattering in the ocean, the atmosphere, and enclosures
- Mathematical and computational models of scattering
- Fidelity assessment for simulation-based training
- Information-theoretic models of auditory detection and spatial hearing
- Scale-model validation studies
By understanding the needs of the end user, whether a tactical-decision aid or real-time training system, ARiA is able to develop models and evaluate their fidelity with scientifically rigorous and tactically meaningfully metrics. Fidelity in modeling is a means to end, not a goal in itself.
Advances in modeling & simulation now enable simulation-based training to augment live training in many applications. But the viability of simulation-based training rests on understanding the relationship between training goals and fidelity and ensuring that the computational load and instructor load resulting from scenario design are optimized jointly with training goals. ARiA is developing new algorithms that automate the process of scenario design that work interactively with instructors to ensure that fidelity is sufficient to meet training goals while balancing restrictions due to computational load.