Army charges on with Joint Simulation System.

The U.S. Army is proceeding with the development of the ground-warfare portion of the Joint Simulation System program, which now is shelved while the Pentagon is searching for alternatives.

WARSIM, the Army simulation program, was supposed to be the land component of JSIMS. JSIMS was designed to be a federation of computerized combat simulations to train commanders and war planners. After years of delays and cost overruns, the program appeared to be getting back on track in late 2002 as it delivered its first version. But in December, the Defense Department pulled the plug on JSIMS and directed the program manager to close the office by September of this year.

According to Col. Kevin Dietrick, the director of Army simulation, the Army made a strong case to the office of the secretary of defense that WARSIM was valuable enough to be further explored. The service has spent $300 million on the project during the past 10 years.

OSD agreed to let the Army "recreate" the program, and restored some of the funding that had been budgeted for fiscal year 2004, said Dietrick.

"We have a significant portion of the land models completed," he said. "We estimate around 70 percent, maybe 75 percent, of the models are done," he told National Defense. "What remains for us now is to make sure that the performance of the system is up to snuff, in other words, that it meets user expectations."

However, only a few months ago, according to one source, WARSIM could not support "even a small, watered-down scenario without crashing for a variety of reasons."

Now, with $14.7 million in the current budget and hopes for considerably more in the coming years, the service is reshaping its simulation program into what it now calls the Army Constructive Training Federation.

ACTF is a collection of simulation models that is going to be fielded in several versions, Dietrick explained. Next year, ACTF starts out with version 1, which will link current simulation systems, such as Corps Battle Simulation (CBS) with Tactical Simulation (TACSIM)--the intelligence piece--and with JCAPS, a joint simulation model at the entity level, said Dietrick. Other simulations also will be added to the federation.

Version 2 is going to bring in a brigade-size exercise from WARSIM.

"We are going to start to see the goodness of WARSIM already in version 2. WARSIM together with OneSAF [One Semi-Automated Forces] will replace the current systems in this second version. ... What we want to do is migrate into what we call the Objective Systems, the Objective ACTF, and that would be made up of WARSIM and OneSAF, plus a system called JDLM, the Joint Deployment Logistics Model," Dietrick said.

WARSIM is the aggregate level simulation that will be used for the higher echelons, to train staffs and commanders. ONESAF is the entity level simulation that will be used for brigade and below. JDLM is the logistics model that will serve both levels, Dietrick explained.

"The challenge for us with ACTF is that we have, especially in the early versions, a fairly sizeable number of models that we are trying to tie together," he said. "As we move towards the objective system--WARSIM and OneSAF--the federation part of it becomes easier."

The current systems will be slowly phased out as WARSIM, ONESAF and JDLM become available.

In building ACTF, the Army must bring in the systems components that JSIMS would have provided, said Dietrick. "We are picking up the pieces off the JSIMS program and we are further advancing them," he said. "They are not satisfactory for the high-end user yet, but we are going to pick them up in a way that it also harvests dollars invested by JSIMS in the past."

The user workstations are some of the most critical components, he said, because they create the "interface between the training audience and the simulation."

"You may have the best simulation in the world, but if the user interface does not work right, the user will be frustrated and unhappy," he said.

The difference between ACTF and the Army's previous approach to JSIMS is that it is not building and delivering one big final package. With WARSIM, "we are going to instead harvest what we have invested to date and deliver it sooner to the user," Dietrick said. "We do not want to do that in one full swoop. ... We want to learn from the user as we develop and deliver smaller pieces of these systems."

The user feedback is going to help the Army make the systems incrementally better, he added. According to Dietrick, it is going to take about five years until the Army reaches an ACTF Objective System capability.

He said that a strong coalition in the Army favors this kind of piecemeal approach. The user community, the Training and Doctrine Command, acquisition officials are the Department of the Army all have given ACTF the green light, according to Dietrick.

While Dietrick said that he was convinced the Army is going to go on with ACTF regardless of what happens with JSIMS, the service hopes that its federation of trainers will be included in the JSIMS alternative. An analysis of alternatives for JSIMS is expected to come out sometime next summer, but it could happen as early as next month.

However, he noted, that there are no guarantees that ACTF will become the land component of the future joint simulation solution. "We could end up out of the AoA [Analysis of Alternatives] with a completely new direction, or an architecture that we have to design," he said. "But we are hopeful that the group conducting the AoA will see the goodness in what we are doing and buy into it as at least the land piece of any joint solution they come up with."

WARSIM is being developed with "the joint perspective in mind," he added. "We think that we could exist independently of the joint solution, although I do not think that that is a wise way to go," he said. "I would really like to see the joint solution for training incorporated into what we are doing because I think it saves the government money and it harvests the value that we are bringing with ACTF."

"We are bringing as much joint flavor in as we can," he said. "If it is not integrated into the joint solution, ACTF will be kept at the Army level."

ACTF will connect to other models. "We will be able to tie into the joint land model and the air models. We are constructing it in such a way that we can pull in any other HLA [high-level architecture] based system."

With JSIMS, the basic problem was that the services had not reached any kind of consensus on how to proceed with the program, Dietrick said. "The joint program was not structured in a way to allow [the services] to be very authoritative."

In ACTF, the Army is proposing to adopt a "minimal number of standards across-the-service solutions and then to allow [the services] to develop their [own] solutions," he said. ACTF will have to have certain features that are common across the Air Force, Navy and Marine Corps. These simulations "have to operate off a common terrain, and you have to link them from a timing perspective," he said.

"With as few requirements as possible imposed, we ought to be able to create, at a higher level, a joint solution that works," he said. "That way you can almost let the services go and build their solutions to satisfy their Title 10 requirements, while at the same time coming together in a closer sort of way than JSIMS did."

Dietrick said that lie hoped the ACTF example would "nudge" the joint community in that direction. "There is a joint aspect to ACTF even by itself," Dietrick stated. "We will be able to tie to the air models and bring them in so that the Army units are not just fighting a ground war, but also have the air war as part of what they are doing. When we train our units even at the brigade level we have to fight in a joint context, we can't any longer just consider a ground unit on a ground unit, we have to do a better job of representing the joint force."

Having a joint force simulation is no longer the sole prerequisite. Another important part is the integration of the live and virtual elements, said Dietrick. "The simulations we build are called constructive simulations, hence [the name] ACTF."

"The virtual pieces are helicopter simulators, armor or Bradley infantry vehicle simulators [for example], where individuals or small crews are trained in a sort of a cockpit environment."

The Army's challenge is to be able to train a brigade at the National Training Center and incorporate that exercise into a bigger event, at the division or corps level.

"That might be run by a constructive simulation," he said. "You want the constructive simulation not just to be able to see what is happening to that one live brigade, but also to stimulate it, to be able to talk to it, it's C4I devices and to cause it to act in certain way, which stays consistent with the plan of that division or that corps."

Then, in a similar fashion, the virtual community--pilots simulating missions in a trainer cockpit--should be linked to the larger exercise.

"This is a critical piece of the future way we will train, and it is a big challenge," Dietrick said. "Our objective systems WARSIM and OneSAF will give us a greater ability to integrate the live virtual and constructive environments."

A C4I adapter--which is built into all Army simulations--allows the constructive simulation to communicate with the C4I devices that are out in the field. The live, virtual and constructive elements are connected through the C4I adapter.

In a brigade, units are stimulated by the constructive simulation by having the division commander write an order through the C4I adapter. That order comes to the unit commander as if it were an order from a higher echelon, having the same map sheet and the same objectives.

"For the current simulations that we have, we have built rudimentary C4I adapters," Dietrick said. "They do an excellent job today, but they do not go far enough." Oftentimes, communication is one-way instead of two-way. "We have the ability to stimulate, but not so much to react to activities with that live brigade," he said. The C4I adapters are "man-power intensive."

To reach the desired level of sophistication, simulations must improve considerably, said Dietrick. Simulations need to adapt to the contemporary operating environment, which encompasses coalitions of varying types and nations, non-combatants, friendly or enemy factions, and complex urban environments.

Many of the current simulations were outfitted in the late '80s to train for the Cold War.

"The current simulations are trying to adapt ... but they really can't," Dietrick said. "In order to do it right, we have to move to the objective systems."

Additionally, "there is a level of reality that is needed and that I do not think CBS provides," he said. "They need a greater sense of realism and that includes the ability to add some of the granularity from OneSAF. WARSIM does not carry it all the way."