BLAST FROM THE PAST: Sandia National Laboratories researchers Candice Cooper, left; Shivonne Haniff, center; and Paul Taylor use a blast impact simulator to study mechanisms behind traumatic brain injury. Most existing simulations developed in stovepipes and are tailored to each community’s requirements. (Photo by Randy Montoya/DVIDS)
Proposed concept for Army migration to modeling and simulation as a service.
by Charles Sanders, Ph.D., and Gene Davis
The U.S. Army established the Army Futures Command (AFC) to address the realization that changes in technology and threats are outpacing their capability development processes. The AFC cross-functional teams are focused on more rapid delivery of new capabilities. However, current Army simulations used to analyze, experiment with and test these new capabilities lack agility and the ability to represent the systems and all domains of the operational environment to support rapid development. In short, modernization of the Army Modeling and Simulation (M&S) Enterprise is required to enable Army modernization.
Current battlespace simulations should be able to provide the required operational assessments and integration into the force structure. However, they are expensive to operate and slow to modify to accommodate the modeling of new systems performance and behaviors. Simulations across the Army are decentralized; they are resourced and managed by six separate communities: analysis; acquisition; experimentation; test and evaluation; and training and intelligence. For example, training simulations are funded through Training and Doctrine Command G-3/5/7 and simulations used by the analysis community are funded through United States Army Deputy Chief of Staff G-8. Most existing simulations developed in stovepipes as standalone capabilities tailored to each community’s requirements. Integration for cross-community cooperation is technically and fiscally challenging. For example, the analysis community typically runs simulations faster than real time to facilitate multiple iterations for statistically valid results while the training community runs their simulations in real time with real players. This means that each community separately creates or collects the same models and data to support their simulation, with limited ability to share or leverage the investments of others.
The long time and high cost to modify a simulation is particularly significant—given the complex, unpredictable and dynamic nature of the forecasted operational environment—the requirement for a simulation to keep up with the pace of change in is a serious challenge. Modifying or updating legacy M&S tools is not sufficient for meeting emerging requirements, as they are inherently man-power intensive, require coding skills and are costly to upgrade and sustain. Therefore, a new approach is necessary to develop and modify simulation capabilities more rapidly and with much less post-development integration re-engineering.
EMERGENCE OF M&S AS A SERVICE
In 2018, the NATO Science and Technology Organization’s Modelling and Simulation Group introduced the M&S as a Service (MSaaS) concept to enable more composable simulation environments that can be deployed and executed on-demand (NATO MSG-136 Report – MSaaS Concept and Reference Architecture Evaluation Report) (see Figure 1). Service oriented simulation is modularly designed, where functions or domains modeled in the simulation are built as individual parts that can be changed or mixed in different ways for the particular use case.The MSaaS paradigm supports agile, rapid, tailor-made simulation solutions from a collection of pre-built and validated models and tools into a unified cloud-based simulation environment whenever the need arises.
MSaaS is intended to promote discovery, reusability and composability of M&S services by one key aspect, the use of modularized simulations. These tools can be easily tailored and composed for a specific use case, rather than relying on monolithic simulations that simply provide all the required simulation functionality or representation, and are slower and more expensive to modify. This approach enables executing each distributed simulation use case with fewer resources and data streams (for more details see, “A Necessary Paradigm Change to Enable Composable Cloud-based MS Services”, Tolk & Mittal).
MSaaS offers an opportunity to better leverage new technologies and tools, plus create inherently interoperable capabilities faster through cooperative and coordinated research and development efforts. Open-source software development, which avoids the limitations of proprietary solutions, enables crowd sourcing which is more effective for innovations. This opens participation by smaller companies with new ideas. Open-source allows for many more experts and more cooperation for more frequent innovations. A good example is the development of the Linux operating system.
MIGRATION FOR ARMY MODERNIZATION
To achieve the required agility and cost limitations, migration from the current monolithic simulation paradigm to one of M&S services that support rapid simulation development is critical for Army modernization to keep pace. However, Army modernization cannot wait 10 to 15 years—the typical timeframe for building a new simulation environment—while the new MSaaS ecosystem is created from scratch.
Therefore, the proposed alternative approach is to integrate cloud-enabled services with existing simulations through a web-enabled interface to provide new modeling capabilities as they develop (see Figure 2). The web-enabled interface will facilitate a service-oriented architecture that provides data sharing between the simulation and each service to provide the desired additional modeling or effect.
The Army Modeling and Simulation Office (AMSO) and the Army Geospatial Center (AGC), are cooperating to establish a rapid integrating environment called M&S Skunkworks, to support integration of M&S services already developed into a persistent simulation framework to test integration of new or updated services as they emerge from both government and industry partners with existing simulations.
The integration environment will include a cloud-enabled interface, called Bifrost that permits simulations to share simulation state using modern commercial approaches and technology under an Army enterprise license. Bifrost will facilitate integrating M&S services as they develop into an ecosystem of current simulations, with access to cloud services for testing before employment in a live cloud service provider. Bifrost’s web-enabled user interface can be used to control multiple sources of simulated entities (simulation agnostic), providing a common single user interface across multiple simulations.
NEED FOR A UNIFIED SIMULATION FRAMEWORK
An ideal scenario for developing future M&S capabilities would be a common modeling framework, in which newly developed technologies and tools from industry, academia and Department of Defense (DOD) laboratories can be integrated and tested, and then retain those technologies and tools in a persistent distributed modeling environment. AMSO views this approach as similar to the Apple App Store architecture, where applications are built to work on the Apple Operating System through specified interfaces. A new sharable synthetic environment can be developed gradually on an established common technical foundation. The advantage of this approach is that this new synthetic environment would provide needed capabilities faster as each tool and or software is developed and tested by scientists and developers; then validated by military operators. This new modeling framework would also enable sharing of investments (cooperative development, where different organizations build separate parts) and facilitate reuse.
Some modeling frameworks have emerged that offer an opportunity for sufficient composability, and to better leverage new technologies and tools, to create inherently interoperable capabilities faster through cooperative and coordinated research and development efforts. These modeling frameworks enable rapid scenario development with the ability to quickly build models of new systems and then play them in the scenario. One of the available modeling frameworks of particular interest to the needs of several AFC cross functional teams is the Advanced Framework for Simulation, Integration and Modeling (AFSIM) (see Figure 3), owned and managed by Air Force Research Laboratory, but shared across DOD and with industry partners.
AFSIM provides the ability to model the capabilities of the participants and to control the interaction of the participants as they move through space and time. The resulting simulations can be:
- Constructive and non-interactive (the user invokes the simulation, which then runs without further interaction), or interactive (the user or other simulation controls some aspects of the simulation).
- Non-real-time (faster or slower depending on the fidelity of the platform component models), or real-time (constrained by some multiple of a real-time clock).
- Event-stepped (simulations proceed according to processing of relevant events) or time-stepped (simulations proceed according to events occurring in succeeding time steps).
Modifying Army simulations for the rapidly changing operational environment and Army modernization focus areas are proving too slow and costly. Therefore, a new approach to composing simulation environments and scenarios is required. The emergence and promise of MSaaS necessitates migrating simulations to this new paradigm. To move more quickly while new simulation frameworks are explored and tested for MSaaS, AMSO and AGC are partnering to establish an M&S Enterprise Skunkworks with a web-enabled interface to facilitate the introduction of new M&S services with existing simulations.
For more information on MSaaS go to: https://csiac.org/articles/a-new-reality-modelling-simulation-as-a-service/#:~:text=M%26S%20as%20a%20Service%20(MSaaS,deployed%20and%20executed%20on%2Ddemand or https://nmsg.sto.nato.int/themes/msaas, or read https://apps.dtic.mil/sti/pdfs/AD1076559.pdf.
CHARLES SANDERS, Ph.D., Trideum Corporation, is a simulation systems architect for AMSO. With over 30 years of experience in modeling and simulation, organization development, strategic policy and planning, training and learning concepts, and has provided technology innovation advice to Office of Secretary of Defense, Joint Chiefs of Staff, U.S. Navy Staff, Joint Forces Command and other executive-level government and industry organizations. He also led studies of emerging technologies for the M&S community, including cloud-based simulation and development of a persistent simulation test bed for coalition experimentation. He previously provided strategy and plans support for the initiation and development of the Department of Defense Training Transformation Program and DOD Advanced Distributed Learning Initiative. He holds a Ph.D. in organizational development with a focus on innovation from Regent University, an M.S. in information systems management from Syracuse University, and a B.S. in chemistry from The Catholic University of America. He is certified as a Navy nuclear propulsion engineer.
GENE DAVIS serves as AMSO’s data standards and acquisition officer. Ongoing projects are AMSO’s M&S Force Structure efforts (Army Organization Server, U.S. Forces integration and migration to the Global Forces Information Management environment, and M&S Real World Threat representation). He has over 15 years of experience in modeling and simulation, live-virtual-constructive integration/event support, and leading technology innovations regarding M&S data ingestion, with experience stemming from the Army National Simulation Center, U.S. Joint Forces Command J7, Joint Staff J7, and the Defense Modeling and Simulation Coordination Office. He holds a B.S. in pre-law and criminal justice from Seton Hall University.