Intelligent Software Agents for Autonomous Intelligence, Surveillance, and Reconnaissance (ISR) Analysis

As net-centric information becomes available to the Warfighter in ever increasing amounts and types, the risk of information overload becomes a serious problem. Tools for filtering, fusing, and analyzing data have become critical components for the effective use of ISR data. The typical strategy for dealing with the large amounts of complex ISR data is to employee additional personnel to present it to the Warfighter, but this is expensive and prone to human error. We have found that the current state-of-the-art in software agent technology should allow the use of software agents to perform the roles of gathering and presenting the information, and in some cases doing so with greater accuracy and efficiency than is possible for human operators. To validate this approach, we will create a prototype of software agents based on a service-oriented architecture (SOA) that can interface with net-centric feeds; analyze, prioritize, and filter the data; and provide it to the operator. The end result will be a set of software agents capable of assisting an operator with Tasking, Processing, Exploitation, and Dissemination (TPED) and Task, Post, Process, and Use (TPPU).