The U.S. Army Science and Technology (S&T) Advisory Group created the Technology Enabled Capability Demonstration (TECD) concept in order to demonstrate and measure progress towards meeting the Army's top ten science and technology challenges. One of the designated challenges is the prevention of tactical surprise at the small unit level. Operating under the premise that soldiers at the squad level lack sufficient situational awareness to prevent tactical surprise, the TECD 3 effort was created to increase small unit situational awareness through the fusion of various planning and intelligence systems into a small unit framework. The Linguistic Geometry Real-time Adversarial Intelligence & Decision-making (LG-RAID) simulation is a lightweight course of action (COA) planning tool that employs innovative algorithms to predict enemy activity in a highly reliable and efficient manner. As such, LG-RAID was selected as a participating application in the TECD 3 federation and was embedded on both individual soldiers and tactical vehicles in a lightweight mission command system. In this paper, we discuss how the LG-RAID simulation improved soldier effectiveness, situational awareness and facilitated the prevention of tactical surprise during the execution of four tactical situational training exercises (STXs) held at Fort Dix, NJ and executed by the Army's Experiment Force (EXFOR). Furthermore, we discuss the integration of LG-RAID into the TECD 3 framework and technical challenges that were overcome. Results of this integration and exercise, presented in this paper, highlight the potential value of embedded simulation at the tactical level.