Is this opinion on personal computer flight simulators too strict or is it actually true?

Question

A couple of days ago I had a discussion with a commercial pilot and his opinion on personal computer flight simulators were a bit surprising to me.

He said that flight simulators, even with expensive payware aircraft are about as close to the real plane as Call of Duty to real warfare.

He said that most of the time the systems and the switches appear to look similar but key factors when operating those systems are missing, mainly because the manufacturer keeps those details more or less undisclosed. The developers don't even know what to simulate and it makes every single flight simulator just a simple game that bares a slight resemblance to reality but that's about it.

This surprised me to say the least, because I thought the well known, premium payware planes simulate the instruments and the systems very well and all the details regarding those systems are available in manuals that we can easily find online.

But I didn't really know that pilot well so I just accepted his answer, I didn't want to seem rude and try to debate him, but I'm very curious if he told the truth or he just tried to exaggerate. I'm especially skeptic about his comment on the undisclosed nature of flight systems.

Answer 1

It isn't true, not as an indictment of the simulator software community as a whole anyway. It may be that your pilot friend doesn't have experience with modern simulators, let's take X-Plane for example...

X-Plane has a "professional use" version, which is simply the regular X-Plane with a commercial license and a few extra features. The underlying software is still the same. X-Plane "for professional use" can be FAA certified as a training device.

This means that the aircraft behaves exactly as would be found in the real world. In fact, X-Plane is sometimes used to verify flight models before more expensive methods are used, and they are relatively accurate. NASA has even written research tools to utilize X-Plane. X-Plane's Garmin G-1000 is incredibly realistic, and X-Plane "for professional use" can drive G-1000 simulator devices.

You can buy other "professional level" simulators for home use like the Redbird Jay which runs the same FAA approved software as the certified simulators. The same is true with Elite Simulator.

Answer 2

No, that is not true.

There are two concepts here: being close to reality and being certified for training. They are not the same thing.

Certified is in essence an insurance policy. It means "sue me if you find my simulation responsible for a crash, because I am so sure it is accurate". That is why it is so expensive. But just because something is not certified to a high standard, does not mean it is not accurate. It just means no body has spent the resources to verify its accuracy because there is no economic benefit of doing so.

---

the manufacturer keeps those details more or less classified so the developers don't even know what to simulate

Well, not when there is an agreement between the developers and the manufacturer. For example Boeing has a relationship with one of the companies in the aircraft simulation industry. The simulation is so detailed that you can execute abnormal checklists using the real Quick Reference Handbook.

And there are things which do not require propriety materials from the manufacturer. On the other end of the spectrum, a Cessna 172 for example, is so simple and so widely used that it is a matter of how many resources one is willing to put in to get the simulation right. There are products which simulate the combustion engine down to cylinder stroke level, and the inner workings of a combustion engine are well understood.

---

flight simulators (...) are about as close to the real plane as Call of Duty to real warfare.

Even PC-based flight simulators can be used in real pilot training. For example it is useful for learning IFR approaches. Surely certain aspects of the simulation are not as real, but that does not prohibit it from being used as a training aid. Again, note that I am saying it is helpful, not that you can log IFR hours by flying in the simulator. And there are simulators which you can log IFR hours on; they are based on the PC-based software which I am talking about.

Let's talk more about certification. There are various levels of certification, covering different aspects of the simulation. For example there are simulators which focus on navigation and instrument procedures, and they are certified by the aviation authorities for such tasks. They are not, however, certified for practicing say an engine bleed failure.

The Level D simulators, which presumably your commercial pilot got his training in, has a very high standard and low tolerance. They are called "Level D" because there are Level A, Level B and Level C simulators. Level A is more relaxed and does not require a motion system for example. Level D is at the top of the game and they are the very best simulator one can get their hands on.

---

every single flightsim just a simple game that bares a slight resemblance to reality but that's about it

You can take the below statement as a personal account or as a bluff on the internet: after my first introductory flight at a flying school, the staff asked if I have heard of PC-based flight simulators and suggested that I learn and practice some basic stuff in the sim while pursuing my PPL. That was before she learnt I would do barrel rolls and hammerheads in the sim...

Answer 3

There is a whole spectrum of flight simulators. These range from mobile apps to room-scale 3D professional simulators. The difference lies in the purpose of the software, i.e. fun vs training. I believe your friend is referring to standard flight simulators such as X-Plane and FSX.

In this case I would agree with him, because it is impossible to simulate every possible failure that can happen in flight. For example, a solder joint can break in flight, causing very strange behavior. Also, the interactions between components internally, for example the XNDR getting the mode S data from the other components will not be simulated, but will be a variable passed to a different module.

Modern avionics is based on Line Replaceable Units, which are sent back to the manufacturer when they fail. This means that the operator of the airplane does not know what is inside in most cases. They do not need to know, because all they do is remove/replace the LRU. The soldering of components is left to the manufacturer.

Next up is the version of the LRU. The developer might have coded a certain model of LRU, which may have changed since the launch of the simulator. Pilots may be flying model 2.3, while the sim still features model 1.1. Model 2.3 may have additional options that model 1.1 does not have, or it may be modified or present itself in a different location in the cockpit. There are fleets in which the avionics package differs from tail to tail, for the same type of aircraft. This configuration issue is very hard to program, because you would have to code every version of a LRU separately.

Answer 4

There are many levels of simulators; and not all are created the same. I can somewhat relate to what the pilot has said.

I used to fly exclusively a very high ranked commercial flight simulator (on my personal computer) with a dedicated yoke and throttle setup. This was for the 737.

Then I had the opportunity to train as a "tour guide" for a flight experience company, which uses Pacific Simulator's equipment.

Thinking "how hard could it be" and feeling confident on my procedures and instrument knowledge, I signed up and then was taken into the cockpit and greeted by this (picture from pacific simulators):

I knew what everything did, but here is what I didn't know and you can't know unless you've sat in a cockpit

1. The fact that the chair moves up and back and not in a straight direction.
2. Three point seatbelts are not fun
3. The brake pedals are difficult to reach if you are short (I'm 5'8")
4. The yoke ... its difficult to hold on to and has some serious resistance.
5. The gear lever, you have to pull out and then lift (or drop it).
6. The trim wheel actually spins
7. Overhead lights can get hot.
8. Ground steering is only on the left side (?!@)
9. Flap lever gates are very rigid.
10. Keeping on the centerline isn't as easy when you have a stereoscopic view.

Many more things such as the level of noise, etc. All this on a static simulator. I can only imagine how wild it would get on a full motion simulator like the ones they have at airline training centers.

Once I got past the entire physical experience - the airplane performed exactly as it did on my home computer - except for one small detail, at full flaps the simulator model tends to rise (due to the increase lift) and you have to compensate by some yoke forward pressure.

On the home PC, maybe it was the autopilot taking care of it, but I could not discern this behavior.

Other than that, and the overwhelming sensation (I was sweating after my first 40 minute flight) its the same...if you can discount the entire dexterity of the simulator experience.

So yes, in many ways your home PC setup is a simple facsimile of the real thing and many of the physical, tactile properties you will never experience unless you sit in an actual cockpit.

As a side effect of this experience, my respect and love for the plane and its pilots increased even more.

Answer 5

First, the ins and outs of avionics are hard. This might be a trivial difference for an amateur trying out an old Cessna 172, but at the other end of the spectrum, if you're a real pilot trying to simulate procedures on something complex like an F-35 or G1000Nxi, you care about lots of little details. Thousands of them. Things like:

• ForeFlight integration
• equipment failure scenarios
• aerodynamics accuracy
• the hundreds of corner cases that pop up

Avionics requirement documents are thousands of pages about how things should transition from on to off, nuances of on-ground vs engines-on vs moving vs weight-on-wheels, maximum limits, mode interactions, message displays, redundancy scenarios when one piece of equipment fails, etc. They're complicated enough that there are corner cases that get missed and heated discussions over seemingly tiny implementation details.

Pilots don't care about these details because pilots are uppity; it's because they're professionals and their lives depend on knowing what to do when the compressor stalls while a different aural warning is going off.

Can you get a reasonably accurate simulation, sufficient to practice things like instrument procedures? Yes. But to get accuracy you need:

• Serious investment and attention to detail (even some paid products have missing features and inaccurate dynamic performance)
• At least a manual, if not technical data, from all the avionics developers for the plane
• Accurate aerodynamics for the plane itself
• A good understanding of performance behavior and issues (e.g. how much does altitude change when you deploy the flaps while the autopilot is set to maintain altitude?)

Additionally the "feel" of flying a plane is missing in home simulators. You can't feel the plane moving, making maneuvers and especially takeoffs and landings very difficult. This was covered extensively in this question on how useful MS Flight Simulator is., and I feel like you care more about the avionics than just the overall simulation usefulness.

Answer 6

If a professional pilot states from experience that they are unrealistic, it must be true: that is the closest reference we'll ever have for the PC-based programs, since they are not normally referenced to actual flight data. Certification of a device means that they have been held to scrutiny and that proof must have been provided at one stage that the behaviour of the system/flight model is realistic. For an uncertified PC-program, who knows? it could be realistic, or not...sales are most important, cool graphics, good sounds etc.

Since Microsoft started to market a PC game called Flight Simulator, there is a bit of confusion about the term. There are:

• Flight simulators, used for training professional pilots. Some devices are used for entertainment purposes as well: you can buy your dad a ride with a certified instructor for father's day.
• PC-based programs, for the general public. These can be used by professionals as well in the comfort of their homes.

The word Flight Simulator exists since almost a century, when Ed Link built his first device for instrument flight training. This became an industry: in flight simulators pilots can be trained in situations that would be dangerous in the air. Emergency situations can be learned by rote and drilled in until all actions are carried out rapidly; an engine fail during take-off can be presented and the appropriate actions can be carried out, without any personal danger.

The industry bears a heavy responsibility: if a crash occurs due to inappropriate pilot actions that were trained in a simulator, the manufacturer of that device faces huge costs, both in money and in human lives. So the aviation authorities demand that the manufacturers demonstrate that what they have made is like the real aeroplane: the response of the simulator must match that of the real aircraft. And that is a big cost driver for simulators for professional use: you must have a recorded response of the real system or aircraft manoeuvre in order to match with the simulator.

But not all tasks need to be trained at the same frequency, and one way of making simulators less expensive is to verify them only for certain tasks. A navigation trainer must have the navigational systems that respond like the real ones - the best solution is to use the original system on board of the aircraft, and simulate the behaviour of the aircraft when it flies on autopilot, which is a calm, controlled flight path without the quirky responses for hand-flown tight manoeuvres. Part Task Trainers have been certified for decades as well.

Image source

The classical distinction between flight simulation and part task system trainers has disappeared in the last two decades: a Flight Simulator is now any program that shows an inside view of a cockpit, some out of the window movement as a function of a joystick input, and lights and buttons that act in a way that seems familiar to pilots. Microsoft said that they are Flight Simulators. Some of the PC programs are used in Part Time Trainers to simulate part of the response of a similar aircraft, and the manufacturers can then claim that they are FAA Certified. It does of course not mean that the reaction of the program is measured against those of the real aircraft, like Full FLight Simulators need to be.

The distinction is still:

• Is the simulator certified. Not the software in general, but the training device.
• For what tasks is it certified. Systems training only, or flight as well?
• If for flight, to what level? Cruise and fair weather, or the full effects of turbulence, windshear, ground effect? Can landings be trained in the simulator? Then we're talking about a Full FLight Simulator, and I can assure you that none of the X-planes, Microsoft games, or FlightGears are powering those.

The commercial pilot you spoke to has had extensive training in professional, full motion, Level D simulators with an incredible amount of proven and certified realism. He compares that with the PC programs called Flight Simulator, some of which are used for some out-of-the-window view for some part task trainer. And he would be correct in stating that the realism of the commercial PC games available for USD 100 is very poor. His opinion is the truth, and how can it not be: a full data package for a commercial airliner that a Level D Full Flight Simulator is certified against, costs between 2 and 10 million USD. X-plane is really not going to spend that.

Answer 7

I understand what your pilot friend is saying and to a certain extent I agree with him.

A desktop flight simulator like FSX, X-Plane, Prepar3D, etc. is a far cry from a level-D full motion simulator manufactured specifically for a particular aircraft with a complete user interface, systems, accurate avionics, etc. But I don’t think it is then fair to say that a desktop PC flight simulator is unrealistic.

I’ve written another post on this subject before but it bears repeating. Desktop flight sims are entertainment products built with a level of realism which allows them to be quickly converted for flight training. They do model basic flight dynamics and other aspects of flying very well, while other areas are lagging, for a number of reasons, including ease of use by neophytes and more concentration on the more commonly used aspects of flight for simming.

All sims have their good points and their limitations. I’ll take FSX as an example.

Positives:

• Economical training program
• Excellent basic flight dynamics
• Very good instrument trainer.
• Allows the user to sample the flight characteristics of numerous kinds of aircraft economically.

Drawbacks:

• Engine dynamics are only realistically tailored for normally aspirated piston engines. It does not model turbocharged piston engines, turboprops, or turbojets well.
• Serious control lag from time of input to response of control surfaces, even with control axes null zones set to zero.
• Systems are modeled at a basic level only and requires additional add-on in payware aircraft to make them realistic. For example FSX’s basic electrical system doesn’t replicate the emergency backup battery system used on their G1000 installation for their high wing single engine line. The Acceleration F/A-18 Hornet is laughably unrealistic.
• Basic avionics flat out suck, including their G1000 system, replicating only a fraction of the features nor doing it accurately compared to the real thing.
• Aging navigation databases make it increasingly unrealistic for use as a training device.
• Air traffic control is limited and somewhat like a human having a conversation with a Barbie Doll.
• Edge of the envelope flight characteristics, including aerobatics and spins are not modeled well and are docile and clunky compared to real life. The FSX Extra 300S is a dog compared to the real aircraft.
• Weather hazards are not modeled correctly and are there only for aesthetic reasons. You can fly an Airtrike into a raging thunderstorm with zero risks aside from getting tossed around a little. Try that in the real world and you’re a dead man.

Many of these problems can be alleviated using add on realistic payware aircraft and weather add-ons. There is also avionics payware which dramatically increase the realism of the avionics used in the game.

I have used FSX for a great deal of my flight training and it has dramatically improved my skills as a pilot. The same mistakes I make in FSX, I will make in the real thing as well!

Answer 8

This is a subjective question which gets to the problem of what does "realistic" mean.

I kind of agree that flight simulators, either in their controls or their flight characteristics, are not really realistic substitutes for flight except for limited purposes. Here are some of the differences:

• In many cases, instruments and controls are complex devices that cannot easily be simulated by a computer. For example, even very simple instruments like altimeters and vertical speed indicators behave much differently in a real plane than in a simulator. Real instruments have weird lag characteristics and respond to gusts and other transient conditions. The real instruments are much more erratic than what you see in a simulator.

• Large commercial aircraft have complex systems system like flight management computers that have large amounts of proprietary software that a PC simulation has no ability to accurately simulate. For example, all FMCs have engine performance models written by the engine manufacturer (GE, Pratt & Whitney, etc). These models are extremely complicated and highly proprietary (secret).

• Simulators do not accurately model the way a plane interacts with the air. For example, glider pilots do not even use simulators like X-Plane at all for this reason. They use specialized simulators like Condor.

• No simulator, either X-plane or Condor, accurately represents gusts and other erratic air movements the way they are in real life. Landing in real life is much more difficult than in a simulator because turbulent air conditions near the ground often require rapid control inputs to land smoothly; none of this is modeled accurately in simulators I have used.

• Simulator controls feel nothing like real controls. Small aircraft use spring loaded yokes and large ones use hydraulics. The plastic rudder pedals you can buy for a simulator feel absolutely nothing like real rudder pedals. Also, the plastic ones lack toe brakes, so the whole braking experience is completely different in a plane than in a simulator.

Answer 9

It will depend of what do you want to learn from your PC Based simulation.

That is also why all type of regulation have different type of simulator qualficiation (ICAO, FAA, EASA, CAAC, ...) If we look at the definition from EASA FSTD(A) 2012 (unfortunately the table from ICAO 9625 Ed4 is better but the document is not available freely on internet):

Appendix 1 to CS FSTD(A).300
Flight Simulation Training Device Standards This Appendix describes the minimum full flight simulator (FFS), flight training device (FTD), flight and navigation procedures trainer (FNPT) and basic instrument training devices (BITD) requirements for qualifying devices to the required qualification levels. Certain requirements included in this CS should be supported with a statement of compliance (SOC) and, in some designated cases, an objective test. The SOC shall describe how the requirement was met. The test results should show that the requirement has been attained. In the following tabular listing of STD standards, statements of compliance are indicated in the compliance column.

Source: EASA CS-FSTD(A) p9 to 25: https://www.easa.europa.eu/sites/default/files/dfu/CS-FSTD(A)%20Initial%20Issue.pdf

That table will help instructors to know which kind of simulator can be used for each kind of exercises. So FFS are not the only kind of simulator used by training centers to teach and to get a Type Rating.

The limitation of a PC Based simulation like X-Plane or FSX, is not necessarily in the representativeness of the airplane but the limitation of information about the aircraft itself. Even with the good aeromodel (like the one in x-plane) you need data to compare the behavior of the aircraft with your aeromodel. It is even possible that the Airplane Manufacturer doesn't extract this kind of data from the Test Flights. The other limitation will reside in the systems themselves. Without having the right documentation how do you want to reproduce the FWS of an A320 even if you have the FCOM has a reference. There is way too much logics behind it to retro-engineer it from a document like that. Same for any avionics, G1000, HNY suite, ....

But the PC Based simulation is not necessary totally useless. For example, to locate the right place of each pushbuttons, levers, switches, ... It can also help (if the system is well simulated enough) to learn the basic of entering a flight plan. Which kind of information are important and so on.

But definitely from a piloting point of view, following the accuracy of the simulation, it can easily start to become negative training if something wrong happen and the behavior from your simulation is not the same as the real avionic.

Negative transfer is defined in the context of this paper as the transfer from one cockpit to another--of different design or configuration--of habits or responses which were appropriate in the former but are inappropriate in the latter, thereby posing a threat to flying safety. This danger has been demonstrated not only experimentally but also in a number of aircraft accident investigation reports. As new aircraft become available to the commercial, military, and private sectors and pilots consequently must transition from older to newer models, the phenomenon of negative transfer becomes increasingly significant. To illustrate the concept of negative transfer and aviation, the author compares the cockpits of two USAF aircraft and how their differences could adversely affect pilot performance. Recommendations are then made on ways organizational flight surgeons can minimize the negative transfer threat to aviation.

Source: https://www.ncbi.nlm.nih.gov/pubmed/7159345