Side Bar

December 2005

Navy Purchases 3-D Simulation For Anti-Submarine Warfare

By Grace Jean

The Navy in July awarded an $11.4 million contract to ManTech Gray Hawk, based in Alexandria, Va., to provide a videogame technology-based simulation for the Naval Sea Systems Command.

According to its designer, Neil Byrne, president of Tactics Unlimited, the 3-D tactical simulator, called “Kill Chain,” has several capabilities, including evaluating mission effectiveness, demonstrating technology, formulating new tactics and training sailors.

NAVSEA’s program executive office for integrated warfare systems is buying the analytical component of the simulation.

“The intent is to make this a multi-mission module, but the initial focus is ASW [anti-submarine warfare],” said Capt. Paul Rosbolt, program manager of ASW at PEO-IWS, which coordinates research, development and procurement of new ASW systems.

The Navy pursued this simulation program at the direction of former chief of naval operations, Adm. Vern Clark, who wanted a capability to interactively model adversary behavior against conceptual systems that the service wanted to pursue.

“As we’re developing new systems, there is concern that if you don’t effectively model what the adversary could do to mitigate the impact of that system, you may end up buying something that is easily defeated by some tactical trick,” said Rosbolt.

The Navy expects the module to deliver in the April-May timeframe, he said.

Byrne, a retired Navy captain and an independent consultant for Symmetron LLC, the Fairfax, Va.-based division of ManTech that is developing the simulation, helped design the Navy’s first PC tactical game, called “NAVTAG,” in the 1980s.

“Comparing ‘NAVTAG’ to “Kill Chain” is like comparing a bicycle to a railroad locomotive,” he said.

“NAVTAG” had about 110,000 lines of source code. So far, the simulations core is finished with more than 400,000 lines of code for “Kill Chain,” he said. Upon completion, the simulation will comprise one million lines of code.

“Kill Chain” programmers have relied on numerous classified sources, including the SHAREM (Ship Anti-Submarine Readiness and Evaluation Measurement) database compiled at the surface warfare development group in Norfolk, Va., to construct the performance characteristics of the ships, said Byrne.

“We are mirroring how well or poorly the fleet does in these kinds of simulations,” said Byrne. “When we’re doing heavy analysis work, we normally use the fleet performance database,” he added.

The PC-based simulation has a man-versus-machine, or artificial intelligence, capability, said Byrne. The AI resides in a Soar engine, developed at Carnegie Mellon University, that uses programmed “decision trees” to decide what to do in a given situation. The engine is used widely in the Defense Department, especially in the Air Force’s simulations, he said. About 80 tactical decision trees are coded into the “Kill Chain” software.

Players can choose to be a tactical action officer or a commanding officer of a ship on the red or blue sides. Or, for analysis work, the simulation can be set to run against itself, said Byrne.

During a recent games conference in Arlington, Va., Byrne displayed a demonstration depicting three U.S. ships in battle against three Chinese ships. The war-fighting could be viewed via a 2-D display, known as the Naval Tactical Data System, and a 3-D display, in which players can watch the ships being attacked or ride along with the missiles being fired.

While the simulation’s graphics may not awe those who have been exposed to commercial war-fighting games, people are struck by the realism found in the simulation, said Byrne. “Things they would expect to occur, actually occur,” he said. “What we want is to be as real as we can.”

The electronics entertainment industry spends roughly 80 percent of its development budget on graphics and 20 percent on realism, said Byrne. To develop “Kill Chain,” the design team flipped the equation, investing 80 percent of its resources in realism.

“You cannot equate us to a video game model because of the enormous amount of realism it takes to make it a viable device,” he said. “The gaming industry doesn’t get down to that level.”

For analysis work, the simulation works without graphics. It’s capable of running thousands of simulations, he said.

“If you play that scenario like you’re really playing, you can spend 45 minutes to an hour and 30 minutes doing this, until one wins. But if you tell artificial intelligence to play itself, it’ll play that game in about two minutes. That’s why we can do thousands of runs,” said Byrne.

The simulation runs on a PC, which is easily reconfigured, he said. “If you want to change a ship, move things around, you can do that, and run them all again,” he said.

Such versatility will allow programmers to adapt scenarios for realistic training, he said.

“How many guys do you think could walk from the bow of a Chinese class destroyer to stern and can talk about all the systems on board? I put a guy in this game, and I make him the CO of that ship, and he’s in battle. By the end of an hour and a half, he’s going to understand what the capabilities in that ship are. And if the next time he plays on the blue side, and he hears that ship on radar, he’s going to remember what’s on that ship. I say, that’s training,” said Byrne.

Rosbolt has also recognized “Kill Chain’s” potential for training sailors.

“We think that it could be turned into a very effective training tool,” he said.

The Navy has a simulation system called “Battle Force Tactical Trainer,” or BFTT, that simulates a realistic battle scenario. The problem with the trainer, said Byrne, is that in order to run the simulation, “you have to have all the enlisted stations manned up to support the decision-making process.” While it’s useful to have all the people talking to each other, it’s a big operation that takes time to set up, he said.

A better alternative would be to train on a simulation like “Kill Chain,” in which multiple players eventually could play against each other or against the computer and gain the same sort of experience.

Future capabilities of “Kill Chain” will include LAN (local area network) and SIPRNET (secret Internet protocol router network) play, with up to 64 stations linked up using a classified system, said Byrne.

The simulation has potential to model for maritime security as well. “Could we do harbor defense? Yeah,” said Byrne.

The simulation began with DD(x) about five years ago, when it funded development of “Kill Chain” as a technology demonstrator. The idea was to model the naval world of today, throw in DD(X) and see how the destroyer would work, said Byrne.

Scenarios included terrorists who came out with rocket launchers in swarming numbers and attacked American ships.

Funding reductions in the DD(X) program ended its sponsorship of the simulation, but the prime contractors of the destroyer picked it up, using “Kill Chain” as the mission effectiveness evaluation tool for the DD(X) preliminary design review in the air and surface warfare mission areas, said Byrne.

In June 2004, the development focus of “Kill Chain” switched from surface warfare and air surface missile defense to anti-submarine warfare. Byrne said the simulation performed 7,000 runs demonstrating the effectiveness of new and projected anti-submarine warfare systems, including Advanced Deployable System, Littoral Combat Ship and submerged underwater vehicles.

During the last few months the developers have ramped up production, said Byrne.

By April, the schedule calls for surface warfare, air surface missile defense and anti-submarine mission areas to deliver, with full capability due in June 2007.

Back To Top