Innovations in live, virtual and constructive (LVC) environments geared for US military joint force training allow a more effective utilization of space and time for training exercises across the globe. As this use becomes more prominent, the need for a suitable after action review (AAR) tool to incorporate an ever-increasing number of data sources is fast becoming a requirement. To perform an AAR fully, data from a variety of input sources must be saved, synchronized, and analyzed. It is important to equip military trainers with an effective tool to facilitate this need for comprehensive data in AARs to maximize the effectiveness of LVC training environments.

Iowa State University is developing an open source software tool for the U.S. Army to address shortcomings of existing AAR tools. Utilizing an innovative modular domain-independent API, users can combine inputs from multiple sources such as simulation data, physiological sensor information, discrete events, and video feeds into a single application. The aggregated information can then be replayed during an AAR session allowing simulation event information to be supplemented with sources not traditionally incorporated in AAR and providing a framework to greatly enhance AAR.

This paper describes such a system (OmniScribe) at its current stage of development, describing its API for the integration of disparate inputs within a single tool and illustrating using a working prototype. It will discuss the current state of the architectural framework, designed to allow users the ability to add additional playback functionality by developing unique modules, and the prototype. Additionally, the paper will briefly discuss the implications a foundation of disparate data stream integration within LVC training will have on future real-time data mining, decision visualization, and enabling deep behavioral analysis of trainee performance.