Hermes

QLI Intelligent Agents and Hermes™ QLI Intelligent Agents

QLI Intelligent Agents

QLI Intelligent Agent Framework

The term Intelligent Agent refers to a software system, which observes, reacts to, and acts upon changes in its operating environment in a rational manner. It is this unique behavior that allows Agents to adapt to and resolve a variety of complex problems.

To allow for the rapid development of Agent based software systems QLI has developed an Agent framework called Hermes. The use of this framework allows for the rapid design and development of light weight small footprint Agent systems.

Rationale and Use of Agents QLI Agents have been applied to a number of different problem spaces and domains in the government arena. These areas range from addressing challenges faced by sailors and Marines (in concert with) the Office of Naval Research (ONR) to the Department of Homeland Security Science and Technology Department.

In support of ONR, QLI has worked with Code 30, Expeditionary Maneuver Warfare and Combating Terrorism Department, in identifying and proactively addressing asymmetric threats in an urban environment QLI has also addressed the problems faced in Code 33, Sea Warfare and Weapons, around challenges brought about by doctrinal changes as a result of Sense and Respond Logistics and Sea Basing which requires new approaches to logistics operations.

Agent Technology QLI Agents QLI Software Agents are composed of five main components; the communications manager, process manager, service manager, and the scheduler form the core of QLI Agents and are required for agent functionality. The fifth component, the application component, is customizable to the problem space and provides the domain specific intelligence that allows agents to execute a variety of diverse tasks.

QLI Multi-Agent Systems The use of Hermes provides a framework for developing and deploying a service-oriented architecture (SOA) built around Agents. It is the SOA approach that allows the use of networks of Agents. Networked Agents or multi-agent systems are able to cooperatively work towards achieving mission goals.

Projects Unified Incident Command and Decision Support (UICDS) System The goal of UICDS is to provide a product to emergency responders that has the ability to manage personnel, direct equipment, and seamlessly communicate, gather, store, redistribute, and secure any mission critical information needed by incident commanders and emergency responders. QLI Agents assist emergency responders by working within UICDS to provide alerting, coordination, and access to new data sources.

Operational Adaptation Background QLI Agent Components • Application – Defines the domain specific knowledge that adds intelligence and provides the interface that allows the user to interact with the agent. • Scheduler – Allows the agent to conduct basic scheduling of tasks, invoked as local services or from other agents. This enables the Agent to prioritize workloads effectively.

Operational Adaptation is an effort being lead by the Office of Naval Research Code 30 to develop technologies that allow the warfighter to increase his operational tempo, impose his will, and retain the initiative against asymmetric threats. QLI has been working with Code 30 to develop Agent based technologies that provide the warfighter with the “right information at the right time” to allow him to gain and maintain situational awareness against these asymmetric threats.

• Service Manager – Registers and invokes services offered by Agents.

QLI is also working with DHS’s Science and Technology Department to introduce Agents into the arena of emergency response through DHS’s Unified Incident Command and Decision-Support (UICDS) System. The UICDS system employees Agents to help manage the complex and highly dynamic environments found in both small and large scale emergency responses.

Author: Merrick Bechini

• Process Manager – Loads, deploys, and runs processes as tasks. • Communications Manager – Provides a generic interface for sending and receiving messages between Agents.

Representations of Agent Behavior Agent Reacting to Environment

Agent Learning from Environment

Sense and Respond Logistics The goal of S&RL is to support the logistics functionality required to deliver enhanced maneuver and fire capabilities to the warfighter. By predicting what will be needed and responding quickly to both anticipated and unanticipated logistic needs S&RL allows the warfighter to maintain combat effectiveness and retain the initiative. In partnership with the Office of Naval Research and BBN, QLI is assisting in the design and development of the Intelligent Agent framework that will provide the system flexibility needed to make S&RL a success.