Observations of the Caravan Guard 89 Exercise

Abstract : Large-scale field exercises, such as those conducted by U.S. and NATO forces in Germany, are both economically and politically costly. One of the more obvious negative aspects of such exercises is maneuver damage -- the damage to civilian crops and property caused by units maneuvering over private and public civilian property during training. Especially during periods of reduced tensions, military planners seek methods of reducing these effects while maintaining force readiness. The use of more simulations and less combat equipment is one way to reduce the maneuver damage and operating costs of an exercise. Simulations also tend to train personnel in certain functional areas (e.g., the deep battle cell of the intelligence staff) better than more traditional training methods, such as field exercises. In 1989, the U.S. Army decided to perform a test to determine the feasibility of employing combat simulations in support of a large-scale field exercise. The first experimental exercise was the U.S. V Corps Caravan Guard 89 (CG 89) exercise held from September 13-20, 1989. The results of this experiment would greatly influence the design and scope of another large exercise less than four months later: Centurion Shield 90 (CS 90), which was part of a larger regular exercise called Return of Forces to Germany (REFORGER). The purpose of this Note is to document RAND's observations of the OG 89 experience and the issues and recommendations that emerged from it. RAND's primary focus during CG 89 was on the simulations and the interfaces between the live soldier-in-the-field portions and the computer-simulated portions of the exercise. A companion document (N-3152-A) presents RAND's observations on the CS 90 exercise. Findings from the CG 89, CS 90, and other exercises are extended and interpreted in a comprehensive report on future large-scale multiechelon exercises (R-4156-A) as part of RAND's project on Unit Training Strategies. (1 table, 2 figures)