ADVANCED TECHNOLOGIES FOR CONTEMPORARY PROBLEMS AND CRITICAL
APPLICATIONS
AGNC is a dynamic organization that strives
for innovations and the infusion of newly emerging technologies within
a variety of systems. Starting as a guidance and navigation company,
our efforts have expanded into multiple complementary areas with the
primary goals of increasing reliability, enhancing autonomy, and
obtaining a better understanding of systems. Critical and complex
systems are the primary target in AGNC's applications. AGNC's
extensive R&D approach blends five areas of work encompassing
guidance, navigation, control, and communications; complex systems
analysis; unmanned systems and robotics; systems health monitoring and
smart sensors; intelligent systems, computer vision, and
neurocomputing
Research and Development Areas
Guidance, Navigation, Control, and Communications (GNCC)
American GNC
Corporation is actively involved in pioneering efforts related to
inertial sensors, interruption-free positioning, and INS/GNSS fusion.
Since its establishment in 1986, AGNC has been actively engaged in the
development of advanced Guidance, Navigation, Control, and
Communications and the automation and integration for autonomous
vehicles, robotics, ground vehicles, aircraft, marine vehicles,
unmanned aerial systems, missiles, spacecraft, and satellites.
Inertial navigation devices comprise the core navigation technology
which is expanded with GPS/INS integration schemes, vision-based
guidance (image processing), advanced filtering, simultaneous
localization and mapping (SLAM), among others. AGNC produced the
world’s first MEMS rate integrating gyroscope in 1999, setting the
stage for continued development of its coremicroŽ IMU product series,
and is among the very first companies to patent
micro-electromechanical (MEMS) Inertial Measurement Unit (IMU)
technology, which is commonly found in most handheld consumer
electronics such as tablets and smartphones.
AGNC also has a rich history in the utilization of advanced control
algorithms (fuzzy logic control, neural network based controllers,
robust and adaptive control) for a variety of vehicle platforms. AGNC
then expanded these GNC efforts to also include the communications
field (hence, GNCC) where current areas of work involve: optimized
routing protocols, mesh networks, low power wireless networks,
cellular networks, high speed data links, etc.
Intelligent Systems, Computer Vision, and Neuroscience
Artificial Intelligence techniques enable the
realization of cognitive systems, where involved methodologies
include: vision systems, stochastic expert systems, Bayesian leaning,
Artificial Neural Network paradigms, cognitive processing, relational
reasoning, planning and scheduling, distributed software
architectures, and agents design. Leveraging high level cognition
functions for aiding the decision support process (such as tactical
decisions in military applications and troubleshooting in the PHM
field) and workflow analysis are key activities at AGNC. In addition,
our vision systems provide machine perception, where advanced digital
image processing techniques are blended with intelligent pattern
recognition methods for highly accurate and real-time analysis.
Everyday vision-based tasks of humans such as
recognizing familiar places, driving a car, or reading another
person’s expression may seem trivial. However, there are significant
technological barriers for implementing such capabilities in computer
vision systems. AGNC is dedicated to developing novel image processing
solutions and has in-depth experience in related areas such as image
enhancement, visual odometry, scene understanding, segmentation, and
target detection & tracking
Due to the current challenges within the
neuroscience field, there are still several areas of opportunity for
new innovations that can have a beneficial impact on system autonomy,
performance, and reliability. AGNC is actively developing technologies
that address the science and engineering aspects of new learning
methods (including architectures, learning theory, analysis of network
dynamics, self-organization, cognitive science, computational
learning, genetic algorithms, and machine learning) for a wide range
of applications (such as image processing, computer vision,
diagnostics, prognostics, control, robotics, optimization, scheduling,
resource allocation, signal processing, forecasting, among others).
Emerging trends for contemporary problems are always in sight for the
integration of cutting edge innovative technologies such as deep
learning applied to cognitive image processing analysis and health
monitoring systems.
Advanced Modeling and Automated Complex Systems Analysis
Complex systems engineering deals with
understanding the sophisticated interrelations among systems,
subsystems, and components, where even simple design changes can have
a significant impact throughout these elements. Advanced modeling that
takes into consideration an element's healthy and degradation effects
combined with stochastic techniques and state-of-the-art health
monitoring for automated analysis together provide a road-map for
addressing complex systems analysis and visualization in a complete
way. As such, AGNC is developing new tools that will improve the
design process by making the interactions among a multitude of
subsystems and components clear and also allow for reducing the
semantic gap among engineers across multiple disciplines.
Unmanned Systems and Robotics
Enhancing system autonomy by infusing
technologies that compile advanced robust control, navigation
architectures, and health monitoring is a key focus at AGNC. Areas of
work include: (a) design of image processing embedded systems and
payloads for drones with minimized size, weight, power consumption and
cost (SWAP-C); (b) robust low level control implementations; (c) robot
virtual reality; (d) advanced system simulation; (e) sensor failure
detection by the analysis of system redundancies; and (f) high level
cognitive functions applied to navigation, mapping, vision,
information abstraction, etc. The AGNC coremicro Robots serve as ideal
platforms for integrating new technologies to reduce operator workload
and increase autonomy and include target detection and tracking,
obstacle avoidance, localization, terrain mapping, navigation and
route formulation, resource allocation, among others. However, the
developed technologies can be applied across a wide range of other
systems as well, where specific applications include: (i) UAV based
surveillance and recognition; (ii) surgical support by miniaturized
robots; (iii) layered sensing architectures for military command and
control; and (iv) terrain analysis and classification by computer
vision.
System Health Monitoring and Smart Sensors
PHM optimizes system
reliability and supports maintenance operations (Condition Based
Maintenance, CBM) as well as automated logistics (depot management and
the supply chain). Depending on the application, PHM realizations
involve: sensing technologies, distributed architectures, system
modeling, failure analysis and characterization, robust pattern
recognition techniques, and regression. AGNC is involved in enhancing
PHM technologies by: (a) blending complex system modeling techniques
with cognitive analysis based on cutting edge Neuroscience and
Artificial Intelligence techniques; (b) integration of innovative
sensor-self-diagnostics schemes for enhanced transducer monitoring,
reliability, and safe operation; (c) developing distributed health
monitoring algorithms within smart sensor networks implementations;
(d) conducting an integral approach where processes at the system
level and sensor level provide system health analysis granularity as
well as sensor data validation and data fusion at different complex
system levels; and (e) developing standardized system health
monitoring frameworks looking to facilitate retrofitting, system
upgrades, and integration. We are committed to advancing the
state-of-the-art in PHM related areas such as structural health
monitoring (SHM) and integrated vehicle health management (IVHM).
For complex
infrastructures where technology continuously evolves, the integration
of configurable and standardized smart sensors with embedded data
acquisition, flexible wired/wireless communications, failure
awareness, self-identification (using Transducer Electronic Data
Sheets), self-learning, and embedded intelligence provide a strategy
to facilitate technological upgrades, increase system reliability,
reduce operator work-load, and increase modularity, scalability, and
extensibility. AGNC research efforts have produced a "Distributed
Intelligent Health Monitoring" framework that consists of Smart
Sensors Networks (SS-Net) serving as distributed computational
platforms that can host intelligent elements (processes) to improve
monitoring operations. For our smart sensors, design for optimized
size, weight, power consumption and cost (SWaP-C) is often a
constraint, such as in aerospace applications. Standalone devices
designed for energy harvesting constraints are also a current need.
Current and ongoing strategic areas of work are focused to: (a)
ruggedization for operation in pressurized, sea, aerospace, and
cryogenic environments; (b) development of non-invasive data and
energy transfer through barriers (aluminum, steel, and composites) by
state-of-the-art ultrasonic techniques to maintain the host
structure’s integrity; (c) standardizing communications and system
architectures; (d) infusing networking capability and the newest smart
technologies (smart-phones, tablets, etc.); and (e) integration of MIL
and NEMA standards.
Areas of work are related to the advancement
of: (i) web-based server containing a database for managing system
health information; (ii) browser-based remote mobile client software
for desktop PCs, laptops, tablets, and smart phones that provides
system health data, information, and knowledge (DIaK); (iii) automated
maintenance guidance to e.g. technicians; (iv) system anomaly
awareness to both clients and main monitoring consoles; (v) secure
networking communications; and (vi) close coupling of smart sensor
networks to enable automated monitoring capability and ubiquitous
system health data visualization. A key focus is to synergistically integrate information
technologies, diagnostics, monitoring and PHM systems, smart mobile
technologies, and secure communications. Target application examples
are:
-
Transportation systems (oil, gas, and water distribution systems)
monitoring
-
Refineries pipelines, tanks, and processing systems monitoring
-
Computerized Maintenance Management Systems (CMMS) -
Complex Systems Maintenance and Repair Guidance -
Logistics and Depot Maintenance
Delivering high
performance computational platforms (in standalone or networked
implementations) that enable real time processing while meeting
additional design constraints is of paramount importance for a variety
of applications including efficient data acquisition, control, and
high level processing. Distributed hardware and software
architectures, the IEEE 1451 smart sensor standards, standalone
systems, and custom hardware design (DSPs, FPGAs, and ultra-low-power
microcontrollers) are baseline capabilities, that when combined with
high end embedded processing devices provide the building blocks for
advanced system realizations that meet the real time processing
requirement. Focus is provided to system designs with optimized SWAP-C
for monitoring applications, smart sensors, drones, advanced
communication bridges, image processing hardware, and computational
platforms for deep learning.
Alliances and Collaborators
Children's National
Medical Center
Georgetown University
Louisiana Tech University
Pennsylvania State University - Applied
Research Laboratory
Rensselaer Polytechnic Institute
The University of Texas at Arlington
University at Albany
University of Kansas
University of Southern California
Virginia Tech
|