As a branch within the Information Technology Division of the U.S. Naval Research Laboratory, NCARAI conducts basic and applied research on Artificial Intelligence (AI) to address problems that are critical to the Navy, Marine Corps, and the broader DOD. Our primary efforts focus on intelligent agents (e.g., integrated cognitive architectures, decision making, natural language understanding), human-machine teaming (e.g., computational cognitive modeling, human-centered computing, human-machine teaming), machine learning (e.g., deep learning), and autonomous systems (e.g., control of distributed unmanned vehicles, sensor-based systems). NCARAI researchers emphasize the linkage of theory and application in demonstration projects and deployable prototypes that employ techniques from AI and related disciplines.
Areas of Research
NCARAI has several research sections, including the Intelligent Systems section, the Adaptive Systems section, the Interactive Systems section, and the Perceptual Systems section. While NCARAI is organized into discrete sections, much of the work is interdisciplinary and scientists across these sections work together on research projects.
- Adaptive Systems - Conducts state-of-the-art basic and applied research in machine learning, autonomous systems, and mobile robotics. Applications include autonomous vehicles (including underwater, surface, ground, and air vehicles, and mobile robots), intelligent decision aids, lessons learned systems, and command and control systems.
- Intelligent Systems - Performs state-of-the-art research in cognitive science, cognitive robotics and human-robot interaction, predicting and preventing procedural errors, the cognition of complex visualizations, interruptions and resumptions, and spatial cognition. The emphasis is on cognitive approaches to enabling more intelligent systems that are able to work more effectively with people.
- Interactive Systems - Develop and enhance computer interfaces for autonomous and intelligent systems, spanning human-computer and human-robot interactions. The group is specifically interested in linking natural language to other modes of computer interaction, such as human gestures, touch-screen and other graphical modes of human-machine interaction. Research includes linguistic analysis of texts and spoken dialog for information retrieval, auditory analysis to address coping with information overload, and phonological analysis of languages and dialects for accent identification. Projects also incorporate human language technology with cognitive modelling and processing.
- Peceptual Systems - Examines issues related to both passive and active sensors required to support autonomous systems. This includes sensors that support autonomous platform navigation, scene interpretation, and teamwork. Techniques include both passive monocular vision, and passive and active stereo and triocular ranging methods, along with algorithms to rapidly interpret the scene. Underlying machine perception techniques are being combined with computational cognitive models to allow for higher-level understanding, and to allow cognitive systems to influence perception, and for perception to prime cognition.