By RICHARD R. BURGESS, Managing Editor

The tactical action officer (TAO), in the hot seat as director of his ship’s sensors and weapons, is alerted on his tactical display screen. A Sunburn cruise missile, launched from a Sovremenny-class guided-missile destroyer, is streaking toward his ship at more than twice the speed of sound.

With only seconds to react, he orders the launch of two SM-2 Standard surface-to-air missiles. The first misses the incoming “vampire,” but the second intercepts and destroys the Sunburn.

Witnessing with satisfaction the ensuing fireball, the TAO orders his ship to launch Harpoon cruise missiles in the direction from which the Sunburn missiles approached. Several minutes later, sensor data — linked via the battle group net — confirms that the Sovremenny was “Harpooned” and is dead in the water.

The scenario above was virtual, not real. The TAO was manning a laptop computer, waging war by responding to realistic sensor data in the comfort of his stateroom, rather than the ship’s combat direction center, using a simulation game called “Kill Chain.” His enemy’s actions and reactions were generated by artificial intelligence.

The TAO has run through this and similar scenarios dozens of times, so many that his actions are second nature and almost reflexive. He is confident that his commanding officer, who will review the scenario, will approve of his performance.

Kill Chain, a software program under development by Symmetron LLC of Fairfax, Va., has attributes that some in the Navy like. Training sessions are replayable and can be recorded, for example. Some service officials say its scenario-based simulations — which harness recent advances in computer gaming technologies — could be valuable in training naval officers in tactics.

However, the future of this and other new computer-based simulation programs in the Navy remains uncertain. Kill Chain is employed by the service as an analytical tool rather than a trainer. Versions of Kill Chain have been used to assess the effectiveness of some of the Navy’s future advanced weapon systems.

Moreover, the Navy appears to be moving gingerly toward the adoption of new computer-based simulation programs, particularly those envisioned as multipurpose tools.

Professor Robert C. Rubel, chairman of the Wargaming Department at the Naval War College, Newport, R.I., noted that the Joint Simulation System (JSIMS), a war-gaming computer program designed to be all things to all players in the joint arena, was terminated last year.

“The future for large joint simulations is bleak. We’ve had the massive failure of JSIMS. With the budget situation the way it is, I don’t see new initiatives in the near future to build a more cosmic simulation,” he said.

Though not envisioned as a joint program, Kill Chain’s supporters do have several uses in mind for it, including war-game modeling and tactical training for individual sailors.

Kill Chain has encountered the programmatic turbulence typical of new technologies adopted by the military. It is proposed to replace the 1980s-vintage Naval Tactical Analysis Game, a board-based software game the Navy began using 15 years ago for officer tactical training that was like an improved version of the toy game “Battleship,” said Capt. Paul K. Rosbolt, program manager for undersea systems in the Naval Sea Systems Command’s Integrated Warfare Systems office. “It actually taught them a fair amount, but that was before the computer age.”

Kill Chain originally was funded in 2002 by the Navy’s DD(X) program as a technology demonstrator software that would model the world’s existing naval warfare situation — with the order of battle of ships, aircraft, weapons and sensors and their capabilities — into which a model of the Navy’s next-generation destroyer could be inserted to demonstrate its capabilities.

“There were very few capabilities in high-end modeling able to measure the warfighting value of stealth in a larger context,” said Joseph Barbano, president of Symmetron and the originator of Kill Chain. “We knew that this capability did not exist. We saw that the Navy had never leveraged the huge advances that had been made in commercial simulation — video game technology. The timing was right to step forward with the kind of tool that would address this capability effectively.”

Because of DD(X) funding reductions, sponsorship of Kill Chain in the DD(X) program ended, but the Navy directed DD(X)’s builder and systems integrator — Northrop Grumman and Raytheon, respectively — to use Kill Chain to evaluate DD(X) in antiair and antisurface warfare roles for the ship’s preliminary design review by Navy procurement officials.

“We were inserting new technologies such as the Close-In Gun System, the Evolved SeaSparrow Missile and new threat systems” to analyze the effectiveness of DD(X), said Thomas P. Boggs, senior scientist for Symmetron. “It’s very easy to enter those technologies and see how those change the outcomes of the scenarios.”

In June 2004, the Navy’s program office for Integrated Warfare Systems assumed sponsorship of Kill Chain, but changed the focus to antisubmarine warfare, reflecting the Navy’s interest in improving its ability to counter enemy submarines. Symmetron was hired to provide an analytical tool that could be used to evaluate the utility of various proposed antisubmarine warfare programs.

Kill Chain is a highly complex Windows-based program designed for commercial-off-the-shelf personal computers. It can be run in real or accelerated time. The game displays a bird’s eye view of a two-dimensional geographic area with tactical symbology, augmented in an inset window, that displays a realistic real-time 3-dimensional graphic view of the action.

The perspective of the 3-D view can be changed at the operator’s will from, for example, the view from any ship in the game to that of a missile heading for a ship. The 3-D view is an optional feature that gives the operator a sense of adrenaline-rushing realism and time-criticality of events difficult to envision only in tactical symbols.

“3-D gives you a more natural view of things,” said James H. McKee, Symmetron’s Kill Chain program manager. “When you look at it from all three standpoints (tabular, 2-D and 3-D), you get a completely different view that allows you to validate that the model is doing [what it is supposed to do].”

“When we first started with 3-D, people viewed it as eye candy, but when you couple the 3-D with the 2-D with the robust data, you’re able to see things much more easily than you would by just looking at a plot,” Barbano said.

For example, some targets on a radar-detection plot are depicted at extremely long ranges. But when trainees flip to 3-D, they see that some of those targets are below the radar horizon.

“The user immediately appreciates the time-space relationship and what’s happening in the environment,” Barbano said. “The kids out in the fleet now are used to seeing multidimensional 2-D/3-D and data representations, and Kill Chain is geared toward those people — sort of a 21st-century user experience.”

“I think were we to develop [Kill Chain as an] operator interface it would be very valuable for TAOs. We don’t have an effective take-it-to-the-ship trainer for the TAO level. I think this would be useful there,” said Rosbolt, a former TAO and destroyer squadron commander. “We think [Kill Chain] brings some unique features in terms of operator-immediate interaction that our other modeling and simulation efforts do not,”

Kill Chain currently is funded only as an analysis tool under a contract for $1.5 million, though training “is an option on the current contract,” Rosbolt said. “We, in the program office, think it has potential,” but thus far the Navy’s “training community has not formalized a requirement.”

For example, Kill Chain could help meet the challenge of providing low-cost, readily available tactical training. Linked together in a network, Kill Chain-equipped computers potentially could train the warfighters of entire battle groups while sitting in port. That kind of broad-based group training now is done with target emulators that send signals from a facility to the ship’s tactical consoles, which requires manning many or all of the ship’s sensor and weapon stations.

One attribute of Kill Chain, Symmetron officials said, is that it can incorporate huge historical databases and detailed reconstructions of results of weapon firings and sensor tests in real environments, such as sonar detections and torpedo firings in the Navy’s fully instrumented Atlantic Underwater Test and Evaluation Center in the Bahamas. The databases enhance realism because models based on them are more accurate than theoretical models. For example, the results of 2,535 missile firings in tests and exercises are incorporated to simulate as best as possible the actual performance of the missiles.

“What you get using this kind of a database is the real kids out in the fleet using the real gear against something that approximates the real threat. Now, you take your new [platform] and throw it in there and see what the outcome is,” said Neil Byrne, president of Tactics Unlimited, a firm specializing in naval modeling and simulation.

The models of the ship classes in the database not only are “anatomically correct,” said Byrne, a retired Navy captain experienced in fleet tactical training who is a consultant to Symmetron. But builders of the database went to the trouble to “find out where all of the key auxiliary [systems] were [on the ships].”

Kill Chain also incorporates the Soar artificial intelligence in use by the Department of Defense. Commonly used in the computer, science, technology and defense industries, Soar, a freeware developed by Carnegie Mellon University, is a cognitive architecture that attempts to obtain an approximation of rational reality, according to the developer’s website.

Kill Chain is 30-percent complete and eventually will embody approximately 1 million lines of programming code, Byrne said. The company is scheduled to deliver antisubmarine warfare and surface warfare assessment capability by April 2006, and full capability by July 2007. The total contract value of Kill Chain for Symmetron currently totals $11.4 million.

At the current stage of development, Kill Chain is usable in a single workstation in a man versus artificial intelligence (AI) scenario, or AI versus AI. Symmetron plans to upgrade the system to allow operation over a local-area computer network and the Department of Defense’s classified Internet.

“Kill Chain has a number of things going for it,” said Rubel. “Certain versions of it incorporate data of actual weapons tests and exercises. So that gives it a bit more credibility. It does some very detailed calculations for damage and for things like detections, engagement, outcomes, missile [versus] missile, gun [versus] missile.

“So, in general, without being able to see the exact codes and algorithms used, it appears as if it has an especially good set of algorithms in it. From a focused naval warfare perspective, it probably can do a range of things, from battle group training to course-of-action analysis to support decision-making.

“The problem with all computer simulations is that it isn’t readily apparent what their algorithms are,” added Rubel, who has seen numerous war game technologies succeed or fail. “It’s all hidden in a computer program.”