The proposed innovation is a framework for small embedded systems that enables quick creation of software components that need guaranteed real-time response for decision making. A small embedded system is defined as one with very limited computing and memory resources. The solution addresses NASA's Automated Exploration Applications of Intelligent Systems (H6.01) program's need for real-time deduction and search algorithms and support for plug and play of automated reasoning components in intelligent systems. Preliminary research on this topic has been performed by the principal investigator using a laboratory testbed. Phase I research is to identify suitable NASA and commercial applications for this technology, evaluate the suitability of the preliminary results to the new applications, identify relevant issues that have not yet been addressed, define a high-level architecture that describes the agent-based framework, and establish the requirements of a visual programming interface needed to assemble components using the framework. Upon successful completion of Phase I, the proposed Phase II work is to build the framework and visual development tool so that it can be distributed as an off-the-shelf product. POTENTIAL COMMERCIAL APPLICATIONS The proposed research will produce a decision making tool that is currently unavailable in the industry. The tool will be marketed in three forms: a) as a strategic element of our component based software design suite b) as an add-in to an existing vendor's integrated design environment product and c) customized to specific classes of applications. Machine health and medical monitoring have been specifically targeted for early marketing efforts. These markets demand highly responsive solutions that are composed of numerous, distributed sensing elements. The resulting tool will be uniquely suited to these classes of applications. Revenues from direct distribution, licensing and application specific customization services are projected to be nearly $3,000,000 in 3 years
