Current Issue

Journal of Advances in Military Studies 2025 KCI Impact Factor : 1.0

pISSN : 2635-5531 / eISSN : 2636-1329
https://journal.kci.go.kr/jams

Current Issue

2026, Vol.9, No.1 All Issues

Combat experiment analysis of common situational understanding operation and weapon recommendation performance in AI command decision support systems

Kim, Deukhwa

, Kim, Jeonghun

, Jun, Jihoon

and 4 other persons | 2026, 9(1) | pp.1~21 | number of Cited : 0

This study empirically examines the application of an AI-enabled decision support system for command and control by integrating AI-based weapon recommendation functions into the Common Operational Picture (COP) procedures of conventional C4I systems. The validation was conducted through battalion-level combat experiments under the Korean Army TIGER framework. The analysis focused on three key aspects: (1) a comparison between AI-generated recommendation time and the time required for human commanders to assess, decide, and issue orders across different target types; (2) a classification of adoption outcomes between AI-recommended weapons and actual fire commands into identical, similar, and non-identical categories; and (3) an evaluation of doctrinal consistency based on simplified tactical concepts (C1–C4) derived from METT-TC. The results indicate that in the initial decision-making phase following target input, human commanders required on average several hundred seconds, whereas the AI system generated recommendations within approximately 1–2 seconds, demonstrating significant time efficiency. Furthermore, AI-generated weapon recommendations showed a predominance of identical or similar adoption across multiple targets, and a high level of alignment was also observed in terms of tactical concept consistency. These findings suggest that AI-based recommendations maintain doctrinal coherence with predefined tactical criteria even under constrained conditions characterized by limited information and manually synchronized operational environments. To be end, our study reflects real-world operational constraints in which sensor and C4I data integration is not fully automated, and COP updates rely on manual input following staff verification and approval. By empirically analyzing AI-supported weapon recommendation and decision-making processes under such conditions, this research provides meaningful practical implications for the deployment of AI-driven decision support systems in operational military environments.