Simulator eases night-vision goggle dangers.

In the surreal world of amplified light, things aren't as they appear. That's why a new simulator for fighter pilots will ease the dangers of training with night-vision goggles.

Instructors said the simulator will enable them to teach pilots on the ground, in an environment where violating safety procedures doesn't have such profound consequences.

"It's totally a safety issue," said Maj. Jonathan Beasley, an instructor pilot with the 56th Training Squadron at Luke Air Force Base, Ariz. "If you look at all the mishaps we've had in fighters using goggles, almost all of them were caused by reliance on some type of visual cue that you're not supposed to be relying on with night-vision devices."

During the initial few rides, pilots often are uncomfortable and disoriented, noted Maj. Jeff Johnson, an instructor pilot with the 310th Fighter Squadron. "It's a different perception that messes with their equilibrium."

Currently undergoing testing at Luke, the night-vision goggle simulator is fitted to existing F-16 flight simulators. It consists of goggles, software, a connecting cable to the flight simulator and cranial movement tracker to record a pilot's helmet motion and to present the proper visual cues. The simulator goggles have the same weight and feel as their cockpit counterparts.

The system is designed to work with all three F-16 flight simulators used at Luke, including the unit training device, re-hosted weapons system trainer, and the networked training center.

The simulator does a good job of visualizing the night sky, asserted Beasley, who also serves as program manager for the networked training center. "If you flip the goggles up and take a look around, it's just as if you were flying around at night, as a typical basic course student does before he goes into night-vision goggles training."

The system is aimed at students who are taking their initial night vision training. During the three to four week course, students fly five training sorties where they practice night formation flying as wingmen. Previously, there was only one night vision simulator--a simple system that only taught them how to don and remove goggles in a dark cockpit.

But actually flying with goggles is an acquired habit. Their most dangerous idiosyncrasies are a lack of peripheral vision and a circular field of view that is limited to only 40 degrees. "You don't have any depth perception," said Beasley. "A light three miles away and a light 50 miles away looks the same through goggles. It's harder to fly formation and figure out how far away you are from the other planes."

Those visual miscues mean pilots must be taught to rely on instrument crosscheck, which mandates constantly scanning their instruments and trusting the data even when their senses tell them otherwise. Johnson, who has used the simulator extensively, said the system greatly helps teach this vital skill.

"What it does best is take away your peripheral vision," Johnson said. "Normally, in the daytime, your peripheral vision sees the horizon and it automatically knows which way is up. When you put the goggles on, it takes that away. Now it's like you're looking through a soda straw. And to take all that information into your brain, you have to move your head around quite a bit. It takes students a few rides to get the hang of it."

Night-vision flight trainees must have a safety pilot with them in case they get disoriented during flight. The simulator enables them to practice more safely and cheaply. "The biggest benefit of the sim is to develop that instrument cross check on the ground when you're not burning time and gas," Johnson said. "When you get the student in the air, his cross check is a lot more efficient."

Proficiency in using night-vision goggles is no luxury. They have become a routine part of night flights, used during most operations except for take off, landing and aerial refueling. Fortunately, the simulator can accommodate both air-to-air and air-to-ground training. "You can get guys in your [simulated] radar scope," Beasley said. "As yon get closer, you can start to see them through the goggles in the simulator."

Beasley said the simulator could actually do a few things that real flight training can't. For example, students in a real night flight will only have a chance to experience using goggles with whatever phase the moon is in that week. The simulator allows instructors to vary the moonlight.

The instructors make clear that simulator training is no substitute for actual flight time with a pair of goggles over your eyes. The simulator isn't a perfect imitation. It doesn't fully reflect the motion of actual flight. Nor does it show the dense air traffic of airliners and small aircraft that are encountered in real airspace.

Perhaps the biggest flaw is that it lacks the processing power to perfectly simulate lighting for goggles that magnify light 8,000 times. "It would take a huge amount of processing power to show that and the shadowing effects 100 percent," Beasley said. But perfection isn't needed. The simulator just needs to be good enough to instill students with the proper procedures for using night vision devices, added Beasley.

The simulator's 2-D graphics and 3-D imagery are powered by Onyx computers from Silicon Graphics. The computers create effects that are reflective (terrain illuminated by moonlight and starlight), emissive (lights, flares and explosions) and 2-D head-up display graphics as seen through goggles, said Brad Morrow, a Silicon Graphics account manager. These three outputs are blended by SensorHost postprocessors from the Air Force Research Lab, which adjusts their relative brightness and also the overall brightness of the scene.

'Training Value' Key to Military Simulations

Computer war games have considerable training potential, but more often than not require significant modifications before they gain acceptance by the military and generate any profits for the developers.

That's the experience of Jim Lunsford, a senior simulations designer for Mak Technologies, in Cambridge, Mass. Lunsford served as an infantry officer and a tactics instructor at the Army's Command and General Staff College.

He is also familiar with commercial off-the-shelf (COTS) software. Based on his experiences as an instructor, Lunsford created "Decisive Action," a division- and corps-level hobby war game that has also been used by the U.S. Army.

Lunsford suggested a path for both government project officers and the designers of COTS software to follow in developing training aids for the military.

Project officers should identify the specific need that can be fulfilled by a COTS war game, he explained. "You don't want to end up having something that doesn't have a real purpose for it," warned Lunsford. Game designers and companies have a natural tendency to claim that their games can be used by the military, but that doesn't guarantee that they have actual training value, he said.

Project officers also need to beware of broad designs that attempt to satisfy so many needs that they satisfy none. As for COTS developers, "always remember you are building a military trainer first," Lunsford says. Any future plans to base an entertainment version off the militarized version must be secondary.

The government project officer must clearly and concisely define the requirements, including the intended training audience, training and learning objectives and the intended training environment. The requirements document should only be a few pages. Anything longer will be so complicated that the project will not be finished on time and within budget, he cautioned.

Mutual understanding is key, Lunsford added. The requirements document should be clear enough to let COTS developers know exactly what they're getting into before they commit to the project. At the same time, developers not accustomed to working with the government need to understand that the process is slow, frustrating and confusing. It can take a long time to get paid, he said, noting that contract deliverables must be submitted on time.

Using the requirements document as a starting point, the project officer and developers need to reach an agreement over what exactly is being created, contract deliverables and the work schedule. Avoid "requirements creep," he stressed.

He said the developer must have sufficient time visiting and meeting with the customer to understand his needs, training environment and culture. The developer should also observe first-hand whatever military training that the simulation is supposed to enhance or replace. Everyone must understand that no major decisions may be made without agreement from all parties, he said.

Developers, Lunsford stressed, should stay focused on the training audience and training objectives and use appropriate fidelity. What doesn't need to be explicit can be abstracted.

"If you're building a sim to train a division commander, then you don't need to model the specifics of a TOW missile," Lunsford said. "If he's not going to be involved in the decision to launch a TOW, then why model it?"

Design the graphic user interface for the military trainee, not for another engineer or a civilian garner, he said. Most military personnel prefer a plain Windows interface to a cool-looking one, said Lunsford.

The project officer and developer must participate in spiral development, he asserted. Create sequential prototypes that are tested by users. Incorporate feedback into the next prototype, Lunsford explained.

He advised developers to remember that at the end of the contract, they must deliver a reliable, working trainer that achieves its intended purpose. If not, everyone has failed. Also understand that it will not be perfect, he said. It's almost a guarantee that more dollars will be needed to enhance and modify the trainer once customers have a chance to use it.

He also recommended that developers budget time and money to train the trainer. Like all tools, a simulation is only as effective as the instructor using it. "Trainers need a bunch of skills sets," Lunsford said. "I normally spend a full day training the trainer."

Trainers don't just need to know how to run the simulation. They also need to understand the concepts behind it. Equally important, they need to know how to handle technical support issues to keep the simulation functioning, he noted.

Lunsford believes that the best results come from COTS projects where the developers and government action officers are familiar with both gaming and gaming technology.

"They know what they want. They have an expectation that's more realistic than someone who doesn't know what a game is about. If someone doesn't have the experience, they depend more on the developer to provide strong input."

Lunsford emphasized that a COTS project include a subject matter expert. "You need a team with a good mix of talent, including a good military expert who also understands gaming technology. If you're developing on your own without adequate support, you focus on the details, and that's what really bogs you down. What helps is finding someone who can help you focus on what is critical," he advised.--MICHAEL PECK