DoD Cloud Innovation, Part 2: Cloud-Enabled Modular Services
Published: May 13, 2014
With sequestration poised to return in fiscal year 2015 and beyond, the DoD is leaning on technology more heavily than ever to mitigate the impact of its declining budget. In no area is this more evident than in joint training, which the Joint Staff is engineering to eventually provide “Simulation-as-a-Service” across the Joint Information Environment.
Last week’s post examined research related to mobile cloudlets as part of cloud computing innovation at the Department of Defense. This week’s post continues the focus on cloud innovation by diving into work the DoD has contracted for cloud-enabled modular services related to expanding use of virtual training solutions.
Multiple trends have driven a shift over the last decade toward greater use of virtual training by the Military Departments. First, evolving technology has provided warfighters with the ability to train in virtual environments using mobile devices, sensors, greater throughout capability, and back-end tremendous computing power for modeling and simulation applications. Second, fiscal necessity has made the use of enterprise technological solutions imperative. Both of these trends should gather strength in fiscal 2015 and beyond.
The Evolution of Joint Training
Way back in 2002, as a result of the exercise Millennium Challenge, a concept for Joint, Live, Virtual, Constructive (JLVC) training emerged at the DoD. This concept evolved over the next decade into the JLVC 2020, a next-generation approach to training that emphasized the use of modular modeling and simulation services hosted in a cloud environment. The idea behind the modular approach was to provide a standardized, flexible, and reusable training solution that promised significant cost reductions across the department. The hosting of modules in a cloud environment further maximized the possibility of reuse beyond the confines of training-specific simulation centers, to include even coalition and NATO partners. In short, all of the Services could be on the same page when it came to the training they experienced, thus enhancing the “joint” nature of contemporary military operations.
Current State and Work Ahead
The road to cloud-enabling JLVC 2020 is a long one that will require a budgetary commitment of approximately $75+ million over the period FY 2014 to FY 2018, according to one estimate from 2012.
Efforts currently underway include:
- Continuing development and refinement of the JLVC 2020 strategy, roadmap, and conceptual design coordinated with the Services, Combatant Commands, coalition partners, agencies, and DOD modeling and simulation community to deliver a future joint training environment reliant on cloud-enabled modular services. Initial capability is expected in FY 2016 and full operational capability in FY 2019.
- Continuing construction of the Joint Training Enterprise Architecture decomposing modeling and simulation, networking, and IT applications into a cloud-enabled modular service supporting Combatant Command and Service joint training requirements.
- Conducting JLVC 2020 Integration Events #2 and #3 to prepare for initial limited operational capability.
This work will be carried out in parallel with the standing up of the DoD’s Joint Information Environment. In fact, the creation of the JIE is a driving force behind the joint training concept as it provides the infrastructure across which cloud-based “Simulation-as-a-Service” will be delivered. DoD budget documents note that the Cloud-Enabled Modular Services for JLVC 2020, or CEMS, for short, will be hosted in the “JIE cloud.” This likely means DISA’s new milCloud capability. However, as DISA continues to certify commercial infrastructure providers, vendors there is always the possibility that the DoD will move JLVC 2020 to a commercially-hosted environment.
Lastly, who’s doing the work providing the CEMS for JLVC 2020? The available evidence points to a single contractor – Roland & Associates – the builder of the Joint Theater Level Simulation (JTLS) capabilities that are to be transitioned into CEMS through reuse of as much JTLS algorithms and parametric data as possible.