| In this research a decision support system is conceptually developed and implemented that can support real-time decision making in complex, uncertain environments. The dissertation proposes that an intelligent decision support system incorporated on a mobile autonomous system can deliver some behaviors necessary for practical applications. Research is conducted to answer the question, “Can an intelligent decision support system improve the performance (threat assessment capability) of an autonomous robotic system and the affected user?” One of the issues for the application of decision support systems to highly autonomous systems is the ability to operate with limited human intervention. Such an approach extends the traditional concept of a decision support system as an interactive system with a human user to a system that exhibits a degree of autonomy. The theoretical foundations for the research are decision making, decision making under stress, decision support systems, econometrics and intelligent agent technology. Single and simultaneous equation econometric models for threat assessment are developed based on expert opinion, electronic sensor data and statistical variable analysis. Elimination of the statistical identification problem provided by the structural equations provides the basis for policy decisions. The models are implemented in a decision support system within SAS Version 8 using intelligent agent technology to increase the autonomy of the system. From an information systems perspective, the proposed concept demonstrates how intelligent decision support systems can be applied to military, and possibly commercial, autonomous robotic systems. It suggests the use of decision support systems as the integrator of multiple information system processing requirements, especially where user intervention is a key to the application. The proposed decision support system provides more autonomy and less user intervention in the area of threat assessment than currently envisioned by existing systems. The proof-of-concept decision support system is evaluated with a multiple criteria model using the analytic hierarchy process to assess both the process of, and outcomes from, decision making and is found to significantly enhance decision making. The methodology is applicable to a wider set of applications that require real-time decision making in complex situations. |
