Virtual airlines
Posted by Ted Pollak on August 6th 2007 | Discuss
Categories:
Tags:
Tray tables up and buckles fastened, folks—thanks for flying as an imaginary passenger in a virtual airplane!
A blogmate of mine is a big flight sim fan, but when he posted a video of himself piloting a large commercial jet from the cockpit (Figure 1) I almost fell asleep—that is until the radio crackled to life and a human air traffic controller (ATC) vectored him for an approach to Dulles Airport in Washington, DC.
People are using Microsoft Flight Simulator in its various iterations on multiplayer servers where, according to Eric Matteson, head of Microsoft Flight Simulations multiplayer division, the record traffic for a single server is 75 human-controlled aircraft in the sky at once, being tracked and vectored by 18 humans monitoring and using virtual air traffic control radar, ground traffic, and communication systems (see Figures 2 and 3). Even more impressive, according to Richard Jenkins, president of VATSIM (Virtual Air Traffic Simulation Network), which is the largest multiplayer flight community, the record across all servers for simultaneous human-controlled aircraft and ATC positions is 1,223. Imagine over a thousand virtual airplanes being flown at once with hundreds of controllers keeping order in the skies. VATSIM has over 130,000 registered members who have logged almost 10 million hours of flight time. ATCs on VATSIM are approaching 2 million hours of controller operations.
I have known about and used multiplayer flight simulations for some time (mostly air combat), but the human ATC part is new to me, as are the incredible complexity and realism of the modern non-combat commercial and recreational multiplayer flight simulators.
My pilot friend Jim, and tens of thousands of other active pilots, partially learn to fly at home on their PCs. Jim credits 10% of his entire piloting skill to Microsoft Flight Simulator, and, based on his experience at Hughes Aircraft in the simulation division, he thinks that commercial and military pilots obtain up to 50% of their skill from dedicated simulators (the ones that actually move and cost hundreds of thousands if not millions of dollars).
When Jim got his pilot’s license he thought it would be easy to become a virtual ATC, but when he applied his application was rejected. As it turns out, the virtual ATC field is very competitive and there is rigorous testing involved to be approved. From the sounds of it some of these folks are training for a career or honing their job skills and their video game play could well be transferable to a résumé, or keep them sharper in the tower. There are four types of controllers that participate in these simulations: ground, tower, approach, and center—just like a real airport.
Over the years, the graphics in these games have come a long way from blocks and lines, and Microsoft’s Flight Simulator X (FSX) is currently the most advanced in graphics and simulation realism. Matt Davis, a flight sim enthusiast, comments on the graphics and experience:
“The graphical elements are fairly extensive in their depth; from ships, AI aircraft, and even airport equipment; the visuals are stunning.
“As you can see from the foggy landings (Figure 4), the visuals from the cockpit are just awesome. There is nothing like riding an ILS beam into an airport in zero visibility. My flight yoke has fingernail markings in it from some of the high crosswind landings I’ve done at night. The visuals of the runways in inclement weather are nice and the rain and snow effects are brilliant. They also tangibly affect landing and take-off friction coefficients. You must de-ice your aircraft in sub-freezing temperatures or suffer some dire consequences. The high-quality graphics allow a pilot flying VFR (visual flight rules) to pick out local landmarks, reducing dependency on the in-flight GPS, and increasing the realism of the flight.”
While the external view eye candy is stunning (Figure 5), almost nothing is overlooked in the attempt to mimic real-life flight from the pilot’s perspective. According to Microsoft’s Matteson, new editions of Flight Simulator will have advanced in-cockpit shadow effects utilizing DX10. When asked about on-windscreen water droplets or fogged instruments, he noted that is possible but not in the cards for the next release. This is where the mod community comes in—the mod community has gone nuts with Flight Simulator, creating all sorts of graphical modifications to increase the realism of the game, the coolest probably being Bug-Splat Pro, which, true to its name, simulates random bugs hitting the cockpit window.
Beyond graphics the options for realism go even deeper. Enthusiasts can file flight plans (required on some servers) and even download real-time weather into the game—allowing pilots to practice in the same conditions they may face later that day behind the yoke of their own airplane.
After learning about all the detail and realism that Microsoft Flight Simulator offers, as well as the amazing multiplayer elements and community of users that embrace this game, I was truly impressed. If we have come this far in 2007—what will the simulations of 2020 look like? One can only look to into the sky and ponder…
Build your own
You too can have a home sim system. Just look at the following setup:
- Pedals: $149
- Civilian” Throttle: $199
- Military Throttle: $149
- Military Sticks: $89–$149
- Yokes: $149–$174
- Multi Function Panel: $189
Flight sim products are available from a number of manufacturers including CH Products (priced above and shown in Figure 7), Thrustmaster, and Saitek.
What do the experts think?
After Ted turned in his article, Jon asked a few of his friends what they thought about it. Here are their (slightly abridged) comments.
From a high-end simulation supplier
It would appear from the article that things have come a long way indeed in the past six years. As a simulation insider, I believe the issue with products like this has always surrounded fidelity of the experience and correct feedback to the pilot/controller. The thing that turns them from a game into a tool is how accurately they interact.
As an example, flight modeling is a real black art. It takes a very specialized engineer (typically an aeronautical engineer with considerable experience) to correctly model all of the flight parameters of a given aircraft type and such engineering skills are prized commodities at companies like CAE, Thales, Boeing, Airbus … and it takes an experienced pilot of such an aircraft type to spot the differences between actual aircraft flight performance and that of a simulation. As such, a considerable amount of tuning is always required to get the simulation to perform the way the test/acceptance pilot wants it to. I cannot comment on how accurate or otherwise the flight model inside a FS is, but I would have to suspect that the amount of tuning required for each aircraft type it simulates has not been performed.
The feedback to the pilot is also key in creating a believable experience, or at least in creating an experience that is deemed to have some kind of real training value. The world of flight simulators makes a serious distinction between the amount of training benefit that one can claim from a given device, and this level of device is governed by a few key factors, and by a whole set of International regulations. This graduated scale dictates that, Cockpit Procedure Trainers at the low end (CPTs) do not need motion systems to give the pilot feedback on movement and do not need visual systems to give the pilot visual cues. They also do not necessarily need high fidelity controls feedback or flight models. Consequently, these devices are considerably simpler and cost less to produce than more high fidelity trainers, but users of them are only able to claim credit for a relatively small part of their training by using such devices. At the high-end you get to Level D Full Flight Simulators (FFSs) and these must have high fidelity controls feedback, flight programs, actual aircraft instrumentation, full 6-Degrees-Of-Freedom (6-DOF) motions systems and wrap-around full daylight display systems. In fact, they are to all intents and purposes supposed to behave like, look like and feel like the actual aircraft. The pilot must be so immersed that he/she is supposed to forget that the device they are flying is bolted to the ground. The real value of such systems is that they allow for the training of circumstances that cannot easily be trained for in real aircraft because they are too dangerous and they lower the operational costs of training significantly for obvious reasons. This all makes them very expensive, but the training credit that can be claimed is considerably higher.
The point is that training credit is given for the immersive training experience and things that are incorrect are considered negative training. In other words, they not only don’t count towards the training value of a device, they actually count materially against it.
I would have to say that I am skeptical that FS is ever going to be more than a really cool game because I am certain that the care and attention required to make a real training product is not being injected by Microsoft. I would add though that I can see where such products are useful in perhaps providing users with an insight into how an aircraft is flown and how airspace is controlled and I am sure that it is a great deal of fun. I would be nervous, though, about claiming that any significant training value can be derived from playing FS.—Phil Cole, Seaweed Systems
From a Simulation builder
The fidelity of the gaming world flight sims is remarkable and they are being widely used in military and commercial flight schools with great success for entry level training. The biggest differences between the MSFS community and the FAA Level C/D commercial and advanced military flight sims comes in several areas—all related to fidelity and adherence to curricula:
- The precision of the synthetic environments: in MS Flight Simulator, much of the terrain is computer generated. For most of what we’re doing today, we use 10m, 5m and 1m sat and aerial photography (we have the entire world), with high-fidelity DTED and precisely modeled air fields and target areas.
- The precision of the visuals: in our case, we’re using WGS-84 datum with 100 nautical mile visibility (or more).
- Correlation to sensors: (IR, NVG, radar) so that what you see in the OTW view matches exactly what you see in a weapon system or radar.
- Correlation to other sims: in Mission Training Center applications, we may have multiple, different simulators flying together, so they all have to have a common operational picture.
- Precision controls: including control loading.
- Curriculum and training scenarios.
- Air Combat doctrine.
In general, it’s simply a matter of fidelity—and for our customers, the fidelity is mandated by FAA Regs or by government specs (like Johnson’s criteria, etc.).
It is interesting however, that in many cases, the perceived fidelity of gaming-based simulations is often higher than that of million dollar sims.—Ross Smith, Quantum3D
From a well-experienced pilot
The other day I celebrated 32 years since my first solo (16th birthday). With over 8,000 hours of flying everything from a Piper Cub (old fabric covered “taildragger”) to large jets; the last time I looked at MS Flight Sim (years ago), it just didn’t do anything for me. Not being a “gamer” or having the time to spend on computer fantasy, it appeared to be just another toy. The author has now compelled me to have another look.
I do get the opportunity to fly my 172 once a week or so. Occasionally, I get a chance to fly five minutes of actual IFR shooting an approach through the fog into the Bay Area. I don’t get the opportunity to fly hard IFR or big iron anymore. The cost of staying current is prohibitive ($1.5k per hour—sim time). Now, I’ll have to go see how realistic MS Flight Sim is compared to the big bucks sims. I’ll let you know…—Robert Praske, anyCOMM
And Ted gets the final word
I would never recommend MS Flight Sim as an “end all” training tool for pilots—but from the feedback I have received; it appears to be a relatively cheap and fun way to learn the basics and can be modded and used multiplayer, for increased realism.
Cole says he is skeptical that any “significant” training is derived from PC flight sims—what is his idea of significant? This is usually expressed in percentage terms. Jim believes he acquired 10% of his knowledge/skill from PC flight sims (especially helpful in learning instrument flight rules, or “IFR”).
To me this is significant; but I fully acknowledge that, in a real-world setting, 10% surely isn’t enough to get you off the ground (pun intended).
