SHINE: RealTime Inference Engine
Minimal Delivery Environment * Realtime Performance * Guaranteed
Originally conceived to meet the rigorous artificial intelligence (AI)
demands of NASA spaceflight programs, the Spacecraft Health Interface
Engine (SHINE) is a unique expert system that serves demanding
realtime environments where embedded knowledge based reasoning can
dramatically impact the performance of a wide variety of applications.
SHINE was built to act as a multimission reusable knowledge based
software tool for monitoring, analysis, and diagnosis of spacecraft
and ground systems through forward and backward inference logic.
SHINE allows systems and applications to easily embrace knowledge
based inference or reasoning logic capabilities into complex systems
Any Sensor or Data fusion intense applications
Advanced Surveillance and Domain Awareness
Onboard Diagnostics and Prognostics
Quality Control Systems
Vehicle Health Management Systems
Its design allows developers to deploy systems that can analyze and
diagnose realtime systems, fielding millions of sensor inputs per
second and making decision/support choices based on expert human
knowledge and streaming data or other peripheral and environmental
The SHINE system uses a Knowledge Base created for your specific
application and offers the ability to reason conclusions based on both
absolute and abstract data. In addition, SHINE allows the creation of
high level abstractions of knowledge which is ideal for Sensor and
Data fusion applications
where realtime reasoning of complex input across multiple domains
becomes a critical factor.
Proven, Efficient and Flexible
Utilizing a highly optimized compiler and a mathematical
representation of knowledge based on data flow analysis, SHINE
maximizes performance and minimizes the amount of processing power
needed in order to deploy a reasoning based system.
SHINE's efficiency means that realtime performance is always
possible, while allowing it to seamlessly integrate with traditional
applications written in conventional
programming languages such as C, C++, or Java.
SHINE offers two modes of operation, a development/prototyping
environment, and a compiled runtime executable mode for creating
“Flight Code” for embedded or distributed applications with zero
Unique RealTime Performance for Weapons Systems, or RealTime
Today, SHINE drives stateoftheart expert systems in realtime
environments. By offering expert system capabilities on an efficient
hardware and software processing level, SHINE allows one to address
realtime problems that were previously considered intractable due to
issues such as:
Software recursion or memory management constraints.
Complexity of integration with external systems.
Difficulty of portability or deployment on limited or
SHINE's optimized inference engine runs without pausing for hardware
system level activities that consume processing power, such as memory
management and garbage collection activities. All necessary storage is
statically preallocated in the final representation state, giving
SHINE a much more predictable execution profile, thereby guaranteeing
performance response times.
Fastest inference engine on the market with realtime performance
capabilities benchmarked at over 330,000,000 rules per second on
Suitable for realtime embedded applications previously deemed
unsuitable for expert systems, such as onboard diagnostics,
sensor fusion, security, and other reasoningbased applications.
Available across multiple platforms, including PCs, Macs, SUN
workstations, and special purpose hardware
The SHINE Advantage
SHINE provides offers a valuable method of building expert systems in
the path of traditional deductive systems.
SHINE can outperform traditional development systems in terms of both
development time and code size.
With multiplatform compatibility, SHINE can crosscompile a knowledge
base into “C” or C++ source code with virtually no runtime support.
The resulting executable size is a direct function of the size of the
knowledge base and not the SHINE development environment. This small
optimized footprint enables deployment on targets with limited memory
or processor resources, making realtime expert system capabilities
available to developers constrained by size, memory, and power.
First expert system designed and developed for hard realtime
Uses minimal resources to allow flexible deployment to constrained
Flexible architecture allows seamless integration with any program
through the use of tight programmatic integration and multiple
For More Information Contact:
Bob Lyons [email protected]
VIASPACE Inc. 171 N. Altadena Ave, Pasadena CA 91107 (800) 5176850