Expert system management systems (ESMS)
Award last edited on: 10/10/2002

Sponsored Program
Awarding Agency
Total Award Amount
Award Phase
Solicitation Topic Code
Principal Investigator
Arthur Gerstenfeld

Company Information


18 Commence Way Suite 4000
Woburn, MA 01801
   (781) 932-5600
Location: Multiple
Congr. District: 05
County: Middlesex

Phase I

Contract Number: F33615-88-C-1720
Start Date: 00/00/00    Completed: 00/00/00
Phase I year
Phase I Amount
The purpose of this project is to show the feasibility of developing an expert system management system (ESMS). The ESMS would interface with the pilot on one side and interface with the other expert systems on the other side. In order to demonstrate the feasibility of the system we have divided the project into three objectives to be performed, consecutively. The first objective is to define the problem. In order to do this we will construct three small expert systems on PC's - one for engine diagnosis, one for route planning, and one for monitoring. Our second objective is to develop a conceptualization model of the expert system management system. This will be based on squadron pilot input, plus inputs from the other expert systems. Our third and final objective is to construct a preliminary expert system management system (about a 20-rule-based system) to show feasibility.

Phase II

Contract Number: 11139
Start Date: 00/00/00    Completed: 00/00/00
Phase II year
Phase II Amount
The purpose of this Phase II project is to go from a feasibility study to a working prototype of an expert system management system. This prototype will be integrated in the intelligent test bed at ,WPAFB. Each of several pilot jobs will have separate expert systems. 'However, to perform each of these jobs the individual expert systems must communicate with one another and interact with the pilot through the expert system management system. The value of the expert system .management system is that it will be possible to very rapidly evaluate huge amounts of data from a large number of subsystems, assess and recognize the current situation, and then consult its .stored knowledge base and rule base to determine alternative courses of action for that situation. The deliverable product will be a 'working prototype that can be used in the intelligent test bed at WPAFB.