- Home
- Applications
- Flight Simulator
Application · Training device manufacturers, operators and integrators
Flight simulator motion platform, delivered with the numbers an evaluator will ask for.
On every other page of this site the customer decides whether the platform is good enough. Here somebody who was not in the room when it was ordered decides, against tests written before the project existed — and then decides again, years later, on the same machine. That changes what has to be delivered.
Expecting us to get your device qualified? Nobody can sell you that. The qualification belongs to the device and to the organisation sponsoring it — the three-party map shows who does what, and exactly where a motion supplier fits into it.
The machine in this photograph will be measured again in five years by people who have never met anyone who built it. That is the design requirement.
What the platform is carrying, and what it has to satisfy
Eight kinds of flight training device, and what each demands of the motion base
The separating question here is not fidelity. It is what the device is being presented against, because that decides which numbers have to exist, who checks them and how often they get checked again.
Airline training
Type-rating and recurrent training simulators
Full flight simulators presented for qualification and then re-evaluated for the working life of the device, often twenty years or more.
Rotorcraft
Helicopter and rotorcraft training devices
A different cue set entirely: rotor-frequency vibration content sits alongside the low-frequency manoeuvre cues and has to reach the crew without shaking the structure apart.
Engineering
Flight test and handling qualities simulators
Devices inside an aircraft development loop, where a pilot’s verdict has to describe the aircraft model rather than the machine reproducing it.
Academies
FNPT and entry-level flight training devices
Bought by academies and schools where the budget is fixed, the footprint is a room somebody already has, and the machine has to survive students.
Crew trainers
Cabin, loadmaster and mission crew trainers
Devices where several people are on the platform, moving around, not strapped into seats — which turns the safety case into the leading design input.
Device type on this list but the qualification basis still open? Tell us who will evaluate it and we will say what changes in the supply →
Three parties, and a motion supplier is none of them
Qualification belongs to the device and to whoever sponsors it, not to whoever built the motion base
The regulations are written around three roles, and it is worth being precise about them because the whole commercial relationship follows from where a motion supplier sits. In the United States that framework is 14 CFR Part 60, which governs a device’s initial and continuing qualification and says nothing at all about who supplied its parts.
The authority evaluates and qualifies. The sponsor holds the qualification, runs the quality system behind it and presents the device for recurrent evaluation. The device manufacturer builds the simulator and puts it in front of the evaluator with a test guide.
We sit inside the third one, as a subsystem supplier. That means the useful question is not whether we can get anything qualified — nobody can sell you that — but whether the motion data the device manufacturer needs will exist, in the right form, before they commit to a number in front of an evaluator.
A supplier who offers to qualify your device is describing work they are not the one doing.
Writing a subcontract and unsure what to put in our scope? Send the device scope and we will draw our boundary on it →
The 3-Party Qualification Map Who owns what, and what a motion supplier can honestly promise each of them
| Party | What they own | What they need from a motion supplier | What we will not do |
|---|---|---|---|
| The authority | The evaluation, and the qualification decision | Nothing directly. They see our data through the device manufacturer’s test guide | Correspond with them on your behalf, or present anything ourselves |
| The sponsor | The qualification, the quality system and every recurrent evaluation | A motion system that still produces the same numbers years later, and a test set their staff can run | Take responsibility for the device’s continued qualification |
| The device manufacturer | The simulator, the test guide and the tolerances proposed in it | Objective test results early enough to write a tolerance they can then meet | Write their test guide, or sign off tolerances on their behalf |
Row three carries a timing trap worth naming. Where the Objective Motion Cueing Test applies, the regulation does not assign the tolerance — the device manufacturer proposes it at initial qualification, and is then held to it. Proposing a number before the motion supplier has produced any data is a guess with consequences, which is why our OMCT results are issued during the build rather than at handover.
What an evaluation actually looks at in a motion system
Six objective tests, and the data has to exist before anybody writes a tolerance around it
A motion base for a training device is not accepted on how it feels. It is accepted on a set of measurements, taken by a defined method, recorded in a test guide and repeated later. Five of the six test the machine. The sixth tests the cue, and it is the one that changed this industry.
The 6-Test Motion Objective Stack What each test measures, and what it proves to somebody who was not there
| Test | What it measures | What it proves | How the number is produced |
|---|---|---|---|
| Frequency response | How the platform answers across the range of input frequencies | That the dynamic behaviour of the machine is characterised rather than assumed | On the assembled platform at the delivered load, not on a bare frame |
| Leg balance | How force is shared between the actuators | That the geometry and the assembly are right, and no actuator is quietly carrying the machine | At defined attitudes, with the load fitted |
| Turn-around check | Behaviour at the instant a commanded motion reverses | That there is no free play or discontinuity where the cue changes sign — which is exactly where a pilot notices one | A commanded reversal, recorded |
| Motion system repeatability | Whether the system still does what it did | That wear, a service visit or a software change has not moved the machine | The same test set, re-run and compared |
| Transport delay | Time from a control input to the motion response | That the cue is not arriving after the picture, which is the order that makes people ill | End to end, one clock, by a method agreed in writing |
| Objective Motion Cueing Test | Gain and phase of the platform’s response, over ten conditions | That the cueing behaviour itself is characterised, not only the hardware underneath it | Ten conditions run on the delivered machine |
Passing all six is necessary and not sufficient, and it is worth saying so. A device can satisfy every hardware test and still be described by line pilots as not feeling like the aircraft, because the first five prove the machine works and none of them describes the cue a person receives. That gap is the reason the sixth test exists.
Writing a test guide and need a tolerance you can actually defend? Ask for OMCT results during the build rather than at handover →
How flight training device work divides
The 4-Configuration Flight Platform Stack
Four configurations, separated by what the device has to satisfy rather than by size. Reading which one a programme is in tells you what documentation exists at handover, which is usually the part that arrives late.
F1
FNPT and entry-level training device platform
Footprint and cost govern · compact cockpit, one or two occupants
Bought by academies and schools against a fixed budget and a room that already exists. It has to survive students, and it has to be serviceable by whoever is on site rather than by us.
Inside the supply
- One delivered unit: moving frame, drives, controller and cabinet
- Loaded acceptance with the objective results issued as a set
- Documentation written for a technician who did not order the machine
- Components on the load path chosen from standard catalogue sizes
Not included: the cockpit shell, avionics and instrument panels, the visual system, and the courseware around the device.
F2
Full flight simulator motion base
Objective test data governs · cockpit, visual system and crew as one load
The configuration built around an evaluation. Everything in it exists so that a device manufacturer can put numbers in front of an evaluator and defend them again in five years.
Inside the supply
- All six motion objective tests run on the delivered machine and issued as data
- OMCT results produced during the build, before a tolerance is proposed
- Transport delay measured end to end by a method fixed in the interface document
- A re-runnable test procedure, with the fixtures and the firmware version recorded
Not included: the cockpit and its avionics, the visual system, the instructor station, the aircraft model, the test guide, and the qualification.
F3
Rotorcraft platform
Vibration content governs · the cue set has two frequency ranges, not one
A helicopter device has to deliver rotor-frequency content on top of the manoeuvre cues. That is a structural problem as much as a control one, and it is decided by the mounting design rather than by the actuators.
Inside the supply
- Structural design that carries the intended vibration content deliberately
- Interface loads issued with the vibration content included, not averaged out
- Service interval set against a duty that includes continuous excitation
- Objective results taken with the vibration content active, not with it disabled
Not included: the vibration generation content itself where it is produced by another system, the cockpit, and the visual system’s own vibration tolerance.
F4
Military, eVTOL and non-standard programmes
The acceptance basis has to be written, because none is published
Where no standard fits — a defence programme with its own regime, or an aircraft type whose own certification is still in progress — the acceptance protocol is drafted jointly before the design is fixed.
Inside the supply
- An acceptance protocol drafted with you, naming every test and its method
- Data issued in a form that can be re-cut against a standard published later
- Component selection reviewed for long-term availability, in writing
- Configuration record detailed enough to re-specify the machine years on
Not included: any regulatory strategy, submissions to an authority, export or security clearance work, and the aircraft programme’s own certification.
F2 is the configuration people under-scope, and it is never the mechanism that is missing. It is the paperwork: the objective results, the method statements, the fixture drawings and the firmware record. Those cost a fraction of the machine and they are the entire reason the machine is worth buying for a qualified device.
Programme sitting between a published standard and a written one? Send the qualification basis and we will say which configuration it is →
Which mechanism sits under the cockpit
The platforms these devices are built on
Flight work pushes harder towards six axes than any other application on this site, for a reason that is easy to state: an aircraft manoeuvre is almost never confined to one plane, and the objective tests measure every axis whether the device uses it or not.
Six axes
6DOF motion platform
The default for flight training device work. All six axes move together, which is what a pitching, rolling, yawing aircraft actually does, and what an objective test set is written to measure.
Hexapod geometry
Stewart platform
Six legs laid out for stiffness and for where the pivot has to sit. In a cockpit the pivot should relate to the crew position rather than to the floor, and geometry decides that far more cleanly than software does.
Three axes
3DOF motion platform
Heave, roll and pitch from a serial mechanism, for devices at the lower end of the range where the acceptance basis does not call for six. Cheaper, smaller, and honest about what it leaves out.
The Axis Read from the Qualification Basis The device standard usually answers this before engineering does
| What the programme has to satisfy | What to specify | Why |
|---|---|---|
| Presenting a device against a published standard that calls for six axes | 6DOF | The requirement is written down. Engineering opinion does not enter into it |
| Reproducing full-envelope manoeuvres, including upset and recovery | 6DOF | Aircraft attitudes at the edge are coupled; three axes cannot carry them honestly |
| Holding an attitude and returning to it, with the pivot at the crew position | Stewart | Where the pivot sits is a geometry decision, and it changes what the crew feel |
| A training device whose acceptance basis does not require six | 3DOF | Three axes that are fully characterised beat six that were bought for the brochure |
The axis argument in general is on 3DOF vs 6DOF, and electric vs hydraulic is worth reading before a training centre purchase — a device running back-to-back sessions in an occupied building is judged on noise, standby power and what the floor looks like after two years, none of which appear on a force curve.
Standard names a requirement but the device scope is still moving? Name the standard and the level and we will return the axis count →
Filmed during testing, not during a demonstration
Three recordings of the unglamorous part
Nobody films objective testing, which is odd, because it is the only footage that answers the question a device manufacturer is actually asking.
Test six, start to finish. The ten conditions driven on a loaded platform, with the gain and phase results appearing as they are computed rather than arriving in a report.
The two tests nobody photographs. A commanded reversal recorded at the actuator, and force sharing measured across the legs at defined attitudes.
Accepted at the real load. The acceptance run with everything aboard — cockpit, visual system and crew — through to the stop sequence and the attitude it settles at.
Prefer to see how a measurement is taken before seeing a machine move? Ask for the objective test procedures and a sample results pack →
The requirement no other application on this site has
The same tests get run again, years later, by people who were never here
A training device is not qualified once. It is presented again, periodically, for its whole working life, and the objective results are compared with the master set recorded at the beginning. That turns a motion platform from a machine into a machine plus an argument that has to survive two decades of staff turnover.
Which changes what is worth paying for. A platform that produces excellent numbers at handover and cannot produce comparable numbers in year twelve has failed at the only thing a qualified device actually needs. Four things decide whether it can, and none of them is performance.
The 4-Part Re-Test Kit What has to be in the crate besides the machine
| Part | What it contains | What it prevents |
|---|---|---|
| Master results | Every objective test result at handover, with the load, the attitude, the ambient conditions and the date attached to each | A future comparison against a figure nobody can reconstruct the conditions for |
| Procedure and fixture | How each measurement was taken, including fixture drawings and instrument requirements | A repeat test that is really a new experiment, producing a difference nobody can explain |
| Firmware and configuration record | Which software version was running for each result, and every change since | Two machines being compared as though they were one, which is what a mid-life update creates |
| Load-path spares list | Motors, drives, encoders and bearings named in standard catalogue sizes, with the ones that affect a measurement flagged | A year-twelve replacement that quietly moves a result nobody expected to move |
This is also the honest reason our components on the load path are deliberately ordinary. Motors, drives, encoders and bearings are specified in standard catalogue sizes rather than in anything clever, because a maintenance team anywhere should be able to source a replacement in year fifteen without needing us to still exist.
Anybody can pass a test on the day it is written. The device you are buying has to pass it again after everyone involved has left.
Sponsor asking what happens at recurrent evaluation? Ask for the re-test kit contents list before you order →
What we need before a quotation means anything
Eight lines, and the first one changes more than the other seven together
Flight enquiries arrive with a payload and a travel figure. Neither of those decides the price of a training device motion base. The qualification basis does, because it decides how much of what we deliver is paperwork.
The 8-Line Flight Device Brief Eight answers, and a flight device quotation becomes arithmetic
| Line | What it decides on the platform | If it is left blank |
|---|---|---|
| The qualification basis: standard, level and authority | The whole documentation and objective test set, which is most of the difference in price | We quote a training device and a qualified one turns out to be needed |
| Gross moving load: cockpit, visual system, seats, crew, cabling | Actuator force, the cue that is reachable, and the travel left over | Sized for a shell and delivered under a complete cockpit |
| The reference point — where the crew sit relative to the interface | Every travel and acceleration figure the programme will quote | Figures stated at the frame, which is not where anybody is sitting |
| Where the cueing filter runs, and who proposes the OMCT tolerance | Controller architecture, and the order in which the test guide gets written | A tolerance proposed before anything has been measured |
| Transport delay target and the method it will be measured by | Command rate, link choice, and whether the figure is achievable at all | A target agreed in a meeting and then measured two different ways |
| Vibration content: is any required, and produced by what | Structural design, and the interface between platform and cockpit | A structure that damps out the cue it was supposed to carry |
| Occupancy: who is on the platform, restrained or moving about | Envelope limiting, stop behaviour and the access arrangement | A machine that cannot be described in the operator’s safety case |
| Service life and support horizon | Component selection, spares strategy and the re-test kit | Selected for today, and re-evaluated for the next twenty years |
Line one is the one that moves the number. A motion base for an unqualified training device and a motion base for a device presented to an authority can be the same mechanism and are not the same supply, because the second one comes with a test set, a method statement, a fixture pack and a firmware record. Nobody should discover that difference after the purchase order.
Basis still being decided between two levels? Send both and we will quote the difference between them rather than one of them →
Agreed before manufacturing, not during commissioning
What the host sends, what comes back, and what happens at the edge of the workspace
Three things get fixed at the interface review and signed by both sides. Each of them turns into an objective test result later, which is the practical reason none of them can be left to be discovered.
Going out
What your host sends
Command set, coordinate convention, the order of the pose terms, units, update rate, and which end of the link the cueing filter sits on. Host link over TCP/UDP, serial or Modbus, with EtherCAT or CANopen below it.
Coming back
What the platform returns
Achieved state, status and any refusal with its reason, at the rate agreed. Enough for a device manufacturer to show an evaluator what the platform actually did, not only what it was asked to do.
At the limit
What happens at the edge
A commanded pose can be unreachable for reasons belonging to no single axis. Each pose and the path between poses is checked before anything moves, and one that fails is refused with a reason returned to the host.
Integrating with a host and a cueing package that already exist? Send their command convention and we will state what changes on our side →
From a qualification basis to a device standing in front of an evaluator
Five steps, and the fourth one happens before anybody writes a tolerance
The order matters more here than on any other application. Objective results produced after a test guide has been written are a comparison; produced before it, they are the basis of one.
The 5-Step Qualification-Support Path What each step produces, and who it is for
ST 01
Read the basis
Standard, level, authority and aircraft type. We say what changes in our scope because of it, and what documentation will exist at handover as a result.
ST 02
Freeze the interface
Command set, convention, units, update rate, where the cueing filter runs, limit behaviour and the transport delay measurement method. Signed by both sides.
ST 03
Build and identify
Manufacture, assembly and geometric identification on the as-built machine, so the controller solves for the platform that exists rather than the nominal one.
ST 04
Run the six
All six objective tests at the delivered load, OMCT included, issued as data while there is still time for a tolerance to be written around them.
ST 05
Support the evaluation
On site or by live link, with the test set, the fixtures and the firmware record handed over so the same measurements can be repeated without us.
Building a programme schedule around an evaluation date? Ask for these five stages with your dates against them →
The seam between a subsystem and a device
What we deliver, what stays with the device manufacturer, and what we decline
Flight programmes go wrong in a specific way: a motion supplier lets a proposal read as though it covers qualification, and the gap is discovered in an evaluation week. The four zones below are written into the proposal, so a gap shows up while it is still an argument about words.
Ours
- The mechanism, controller, drives and cabinet
- Motion produced at whatever command level the signed interface document sets
- A hard-wired stop circuit, mechanical limits and workspace protection
- All six objective test results, taken at the delivered load
- The re-test kit: procedures, fixtures, firmware record and spares list
Never ours
- The cockpit, avionics, instruments and visual system
- The aircraft model and its flight dynamics
- The test guide, and the tolerances proposed inside it
- Qualification, and any correspondence with an authority
- The training, and whether the device delivers any
Outside our class of machine
- Complete simulators supplied turnkey
- Certified seats, harnesses and restraint systems
- Image generation, projection and dome systems
- Regulatory or certification consultancy of any kind
Signed by both
- The qualification basis the platform is built against
- Where the crew reference point sits, and the envelope written at it
- The transport delay figure, and the method that will produce it
- What the re-test kit will contain, item by item
Being named as a subsystem in somebody else’s submission is a better outcome than being described as something we are not in our own.
Drafting a subcontract scope for a motion subsystem? Send the draft and we will mark where our supply ends →
Where to go next
Reference reading for a flight device specification
One level up, one sibling with the same physics and no authority attached, the axis question, and the words a specification needs to use.
Parent application
Motion platforms for simulators
The general case underneath this one: what a person can detect, what tilt coordination is really doing, and how late the whole chain may be.
Sibling
Driving simulator motion platforms
Same physics, no authority at the end of it — and a horizon close enough that tilt coordination creates a second problem.
Axis decision
Three axes or six, and what each leaves out
What three axes cannot reproduce, what six cost in moving mass, and why flight work sits almost entirely on one side of that line.
Vocabulary
Motion platform glossary
Gross moving load, motion reference point, washout, tilt coordination, transport delay — one agreed definition each, so a test guide and a purchase order describe the same machine.
Three outside documents are worth having open while a flight device specification is being written, and none of them is ours. 14 CFR Part 60 governs the initial and continuing qualification of flight simulation training devices in the United States. FAA guidance bulletin NSP GB 16-03 sets out how the Objective Motion Cueing Test is implemented, including the point that Part 60 assigns no tolerance of its own. ICAO Doc 9625 is the international manual of criteria for qualifying flight simulation training devices, and its fidelity boundaries are what many OMCT programmes work to.
Asked before a specification is written
Flight simulator motion platform FAQ
Can you get our simulator qualified?
No, and nobody selling you a motion base can. Qualification belongs to the device and to the organisation sponsoring it; 14 CFR Part 60 governs the qualification of the device rather than the supply of its components. What a motion supplier can do is deliver a subsystem whose objective test results exist before the device is presented — frequency response, leg balance, turn-around, repeatability, transport delay and the OMCT — written so your test guide can quote them. A supplier who offers to get a device qualified is describing work they are not the one doing.
Do you supply the cockpit, avionics or visual system?
No. We supply the motion subsystem: mechanism, controller, drives, cabinet, the interface and the records describing them. The cockpit shell, instruments, avionics, visual system, instructor station and the aircraft model all need knowledge of the aircraft rather than of the machine producing motion, and they stay with the device manufacturer. Where a customer wants one supplier for a whole device, we are the wrong company and it is cheaper to establish that in week one.
Which objective tests do you run, and when?
Six, all on the delivered machine at the delivered load rather than on a representative one: frequency response, leg balance, turn-around check, motion system repeatability, transport delay and the Objective Motion Cueing Test. The timing matters as much as the list — OMCT results are issued during the build, not at handover, because the tolerance in your test guide is proposed by you and you should be proposing it around a measurement rather than around an expectation.
Who decides the OMCT tolerance, and what if the platform misses it later?
The device manufacturer proposes it at initial qualification. FAA guidance is explicit that Part 60 does not assign a tolerance of its own, and ICAO Doc 9625 Edition 4 provides fidelity boundaries many programmes work to. That has a practical consequence worth planning around: whatever number goes into your test guide is the number your device is measured against for its working life, so proposing it before any data exists puts a programme in front of a figure it may not be able to hold. Our results come first for exactly this reason.
Can you put a motion base under a cockpit we already have?
Usually, and three things get measured before we quote rather than after. The gross moving load of the cockpit as it stands, including the visual system, seats, crew and cabling. The centre of gravity and its height above the proposed interface, because a tall load spends travel on tilt. And the mounting structure of the existing cockpit, which was designed for whatever it sat on before. One thing to be clear about afterwards: the objective results then describe the combination, not the platform alone, so a later change to the cockpit is a change to the tested article.
The device will be re-evaluated for twenty years. What happens then?
Four things ship with the platform so that a later comparison is meaningful: the master objective results with their conditions attached, the test procedures and fixture drawings, the firmware and configuration record, and a spares list naming the load-path components that can move a measurement. What we will not claim is that the machine passes in year twelve — wear, environment and utilisation belong to the operator, and a supplier guaranteeing a measurement two decades out is guaranteeing something they will not be present for. What the kit gives you is attribution: if a number moves, you can say which part moved.
Do you take export-controlled or security-cleared programmes?
No. Work requiring export-control handling, security clearance or national-security caveats is outside what we can support, and where a defence programme carries those obligations we are the wrong supplier. We would rather say that in the first reply than three months into a procurement. Military training programmes without those obligations we do quote, and they usually sit in configuration F4 because the acceptance regime is written rather than published.
Why is your quotation higher than a supplier offering the same payload and travel?
Ask what is in the price besides the machine. Whether all six objective tests are run and at what load — on the delivered platform or a representative one. Whether the OMCT is included and when the results arrive relative to your test guide. How transport delay will be measured and by whose method. Whether the re-test kit, the fixtures and the firmware record are included or quoted later as documentation support. Those four account for most of the gaps we see between quotations that look comparable. Where a competing quotation genuinely covers the same scope for less, we would rather hear about it than guess.
Commercial questions — lead time, payment terms, spares, installation — are answered in the full FAQ.
Send the qualification basis. We will tell you what the supply has to contain.
The standard and level the device is being presented against, the load including the visual system and crew, and where the crew sit are enough for a first pass. Where the programme actually needs a whole simulator, or carries export obligations outside what we can handle, you will be told that in the first reply.
What comes back first
- What the basis you named changes in our scope, and what documentation exists at handover because of it
- Which of the four configurations the programme is, and the axis count the basis implies
- Which objective tests we will run, at what point in the build, and what the results pack contains
- A delay budget with our part filled in, and a measurement method to argue about