Suitability of Agent Technology for Military Command and Control in the Future Combat System Environment

Abstract

The U.S. Army is faced with the challenge of dramatically improving its war fighting capability through advanced technologies. Any new technology must provide significant improvement over existing technologies, yet be reliable enough to provide a fielded system. The focus of this paper is to assess the novelty and maturity of agent technology for use in the Future Combat System (FCS). The FCS concept represents the U.S. Army's ''mounted'' form of the Objective Force. This concept of vehicles, communications, and weaponry is viewed as a ''system of systems'' which includes net-centric command and control (C{sup 2}) capabilities. This networked C{sup 2} is an important transformation from the historically centralized, or platform-based, C{sup 2} function since a centralized command architecture may become a decision-making and execution bottleneck, particularly as the pace of war accelerates. A mechanism to ensure an effective network-centric C{sup 2} capacity (combining intelligence gathering and analysis available at lower levels in the military hierarchy) is needed. Achieving a networked C{sup 2} capability will require breakthroughs in current software technology. Many have proposed the use of agent technology as a potential solution. Agents are an emerging technology, and it is not yet clear whether it is suitable for addressing the networked C{sup 2} challenge, particularly in satisfying battlespace scalability, mobility, and security expectations. We have developed a set of software requirements for FCS based on military requirements for this system. We have then evaluated these software requirements against current computer science technology. This analysis provides a set of limitations in the current technology when applied to the FCS challenge. Agent technology is compared against this set of limitations to provide a means of assessing the novelty of agent technology in an FCS environment. From this analysis we find that existing technologies will not likely be sufficient to meet the networked C{sup 2} requirements of FCS due to limitations in scalability, mobility, and security. Agent technology provides a number of advantages in these areas, mainly through much stronger messaging and coordination models. These models theoretically allow for significant improvements in many areas, including scalability, mobility, and security. However, the demonstration of such capabilities in an FCS environment does not currently exist, although a number of strong agent-based systems have been deployed in related areas. Additionally, there are challenges in FCS that neither current technology nor agent technology are particularly well suited, such as information fusion and decision support. In summary, we believe that agent technology has the capability to support most of the networked C{sup 2} requirements of FCS. However, we would recommend proof of principle experiments to verify the theoretical advantages of this technology in an FCS environment.