Intelligent Systems
Critical Infrastructure Protection
Intelligent Control of Mobility Systems
Intelligent Manufacturing Systems
Intelligent Open Architecture Control of Manufacturing Systems
Research and Engineering of Intelligent Systems
Intelligent Systems
Division Contact: Al Wavering
Critical Infrastructure Protection
The computer systems that control industrial production and distribution
have been designed first and foremost to meet performance, reliability,
safety, and flexibility requirements. Yet these systems increasingly incorporate
connectivity and remote access capabilities. Industry has begun to appreciate
that increased connectivity and openness are introducing serious vulnerabilities
into their operational systems. One critical problem of immediate concern
is the absence of methods to specify and verify the security characteristics
of control system components and networks. These are problems that the
Common Criteria for Information Technology (IT) Security Evaluation, developed
by NIST and the National Security Agency (NSA), are intended to address.
To develop IT security requirements for industrial control systems, NIST is working with industry through its leadership of the Process Control Security Requirements Forum (PCSRF).
The goal of the program is to increase the security of computer systems that control production and distribution in critical infrastructure industries, including electric power, oil and gas, water, chemicals, pharmaceuticals, metals and mining, pulp and paper, and durable goods manufacturing by defining and applying standard information security requirements, developing information security best practice guidelines and conducting outreach activities, and developing laboratory and field test methods for information security products and approaches applied to the process control sector.
Contact: Fred Proctor
return to top of pageIntelligent Control of Mobility Systems
Intelligent vehicles contain a variety of interdependent subsystems. Specific subsystems may be assigned to handle such tasks as recognizing obstacles and landmarks, mapping the environment for navigation, detecting and reacting to unexpected objects, locating objects for manipulation, comparing scenes for security and surveillance, and planing paths. Open-systems architecture standards are needed to promote efficient integration of these and numerous other subsystems and to simplify development of intelligent mobile systems for government, industrial, and consumer purposes.
In this collaborative project, researchers are addressing three categories of measurement and standards needs: interoperability; real-time sensing for control; and evaluation of component and system performance.
The Department of Defense (DoD) initiated plans for deployment of robotic vehicle platforms in the battlefield. It intends to standardize the platform architecture and interfaces, thereby encouraging use of commercially available, "plug-and-play" components and promoting reusability and interoperability on a variety of ground vehicles. NIST contributes to several DoD initiatives that support these objectives.
We use our Real-time Control System architecture to provide a systematic analysis, design, architectural framework, and implementation methodology for developing open and real-time sensor-based control systems. Functional task execution is viewed hierarchically with motor skill functions, such as steering and speed control, performed at lower levels and coordinated actions between vehicles performed at higher levels. The control system uses sensory information to guide the intelligent vehicle in the execution of complex tasks. Planning for task execution and for adaptation to changes in the environment are also parts of the total hierarchy.
We also use real-time sensing and perception as keys to effective control. Accelerometers, inertial navigation systems, and differential global positioning system receivers measure vehicle motion through the environment and precisely locate vehicles, targets, obstacles, and terrain features on a map database. Vision and laser range imaging, enabled by stationary and moving sensors mounted on the vehicle, identify dynamic changes in the environment and provide detailed local terrain and obstacle information during navigation. We carry out research in real-time measurement for control—particularly for imaging sensors—and develop performance measures for evaluating the performance of these systems.
A good example is our work in support of the Department of Transportation's Intelligent Transportation Systems program. For the National Highway Traffic Safety Administration, NIST developed a real-time measurement and roadway calibration system for evaluating the effectiveness of on-vehicle crash-avoidance systems.
Contact: Maris Juberts
Intelligent
Open Architecture Control of Manufacturing Systems
Over the past two decades, information technology has dramatically increased the intelligence of the upper levels of manufacturing systems. In the next 20 years, this intelligence will reach down to the factory floor. Individual machines will become much smarter. They also be able to communicate more broadly, integrate more easily, predict results, avoid or diagnose mistakes, use extensive in-process gauging, and use scientific models to optimize productivity.
Industrial automation users are paying billions of dollars each year to integrate, customize, maintain, and train staff to use proprietary equipment. Open architecture control—based on a common set of well-defined components and interfaces—can help reduce life-cycle costs, improve quality, and increase productivity by simplifying these functions. To get there, user groups are pressing for standardization of controller functions. NIST is working with industry groups in three key industrial segments—machine tools, automated dimensional measurement equipment, and robotics—to bring users and vendors together to work toward standardization and, ultimately, desired levels of interoperability.
This program has established testbeds that are helping industry validate architecture and interface standards needed to get open architecture controls for machine tools, robots, and automated metrology equipment into the marketplace. Researchers also are developing conformance tests that companies can use to make sure that the systems they buy work together. Through this program, NIST aims to reduce life-cycle costs of automated manufacturing equipment by reducing the effort needed to integrate controllers and, at the same time, accelerating the implementation and commercial availability of controllers with advanced capabilities that improve manufacturing equipment productivity and product quality.
Contact: Frederick M. Proctor
Research
and Engineering of Intelligent Systems
Intelligent control
is a critical enabler for realizing increases in manufacturing productivity
and quality. Industry roadmaps call for a variety of intelligent control
capabilities, including self-diagnostics, adaptive control, error compensation,
thermal and load compensation, and tool wear and breakage monitoring.
All of these functional advances in machine control require real-time
measurement capability and real-time models of the machine, the part,
and the process itself. Today's machine controls, in contrast, are limited
to acting on symbolic commands and producing repetitive motions.
There is a need in the intelligent control research community for quantitative,
objective performance metrics to define and structure the problem domain,
to improve communication among researchers, and to lay the groundwork
for methods to describe and evaluate future new products resulting from
research. Measurement and standard needs include performance metrics for
components and systems, data standards and tools for knowledge engineering,
and an understanding of how to incorporate learning and adaptive capabilities
in open architecture machine controls.
In response to these needs, the researchers in this project area are addressing issues in three key areas:
 |
Metrics. A major program objective is to help to define quantitative metrics and methods for assessing the performance of a system or its sub-components and to facilitate field-wide adoption of these metrics. |
|
Knowledge Engineering. This program is developing approaches for building real-time models in an open systems environment and applying these approaches in a variety of prototype systems in different domains. |
|
Learning. In this program, real-time control architectures, developed at NIST and elsewhere, are being extended to include learning, self-optimization, self-diagnosis, and adaptive control. The aim is to assist industry in developing principles, metrics, and standards pertaining to these technologies. |
Accomplishing these objectives will establish a clearly defined framework within which intelligent system technologies can be readily measured and integrated by U.S. manufacturers.
Contact: Elena
Messina
Date
created:
December 17, 2001
Last modified:
Aug. 07, 2007
Contact: inquiries@nist.gov