The use of games for training is increasing in the Defense community; however, this type of training can ignore the best interests of trainees and trainers. DARPA's DARWARS program sought to address this deficiency by combining competency-driven access to game-based team training with the re-use of training content. For example, DARWARS Ambush!, a game-based training environment developed for DARWARS, allows warfighters to create scenarios that re-create their battlefield experience. Then, using DARWARS graphical tools, warfighters tag scenarios using meaningful search terms, and publish the scenarios to a registry. Instructors search the registry for scenarios that address their unit's training requirements. Once selected, a scenario is automatically integrated with an appropriate Shareable Content Object Reference Model (SCORM) course. Then the DARWARS delivers both the course and the training scenario. Assessments from the training session are reported to the DARWARS Learning Management System (LMS), which uses the assessments to direct trainees to appropriate new or remedial training.

This paper describes how we integrated the following technological advancements in an end-to-end demonstration for the benefit of the warfighter training community: (a) The DARWARS Ambush! convoy trainer, developed under DARPA sponsorship, provides PC-based, simulation-based training for small teams. Ambush! is in use at multiple military installations, and has trained over 30,000 warfighters. (b) The Advanced Distributed Learning Registry (ADL-R) is a searchable index of content metadata that can be resolved to content in distributed repositories. The primary purpose of the ADL-R is to support searching for, discovering, and re-using learning content. (c) A framework for simulation-based training integrated with courseware targeting individuals and small teams. Assessments from the simulation-based training are reported back to a SCORM-conformant LMS, where the results are used to direct each team member to new or remedial training. The Joint Advanced Distributed Learning CoLab (JADL) sponsored this work.