Universal Motion Simulator: Real Enough to Evoke Panic

Interesting, and certainly HAS real-life application, but when you watch the video, it looks like the "vehicle's" axes of rotation are a good ways behind the pilot's seat, and not consistent with each other (necessary due to the points of attachment/rotation, and the extremely large and robust joints involved), which would seem to cause the experience to miss a very important part of the "seat-of-the-pants" experience simulating real-life flight. The issue is that rotation as felt by the pilot will be at different distances from the pilot's sensed axis of rotation, with each plane of rotation having its own center, and no consistency from plane to plane.

To see this issue in action, try driving a school bus on a twisting road (or ride right behind the driver, if you don't drive buses), then take the same trip sitting just in front of the back axle(s) of the same bus, with the same driver. Since you are seated in different places relative to the bus's center of rotation (and ITS axes of rotation are all pretty much centered in one place) the change in motion your stomach and ears feel is changed with seating position. But at least they ARE consistent, sharing a common center. AND they are only robust in two dimensions, since the bus isn't flying (depends on how wild the driver is, I suppose, but you only fly once on a bus with a driver THAT wild, no?).

Now imagine the same effect, but with the three axes of motion at different distances from you, AND from each other.

I suspect that the best simulator ever existing, tethered to a solid, gravity anchored structure, will always suffer these deficiencies, but I think the "real life"-likeness of this experience is somewhat over-stated, as a result.

Simulation (especially flight simulation involving a motion platform and visual system) generally will require a number of trade-offs.

With a motion system one of the trade-offs is between a realistic acceleration cue and negative cueing. Ideally, the simulation should try to give as close to real life an acceleration cue as possible. Since the simulator is attached to the ground with limited stroke actuators or fixed length rotating arms, that's not possible, at least not for any significant length of time. Very large systems such as NASA's Vertical Motion Simulator (VMS) (check it out here), or Toyota's driving simulator (seen here) are able to provide longer sustained acceleration cues, but those come at a cost (space, dollars yen).

Back to the topic of a trade-offs.....a motion system may be able to provide a very realistic cue, but once the actuators reach their limit (or get close), the cue needs to be washed out (faded out and the platform brought back to a neutral point). The acceleration has to not only stop, but the built up velocity has to be brought to zero (thus requiring an acceleration in the opposite direction), and the excursion of the simulator also returned (requiring velocity in the opposite direction). Such motions must be done at a level low enough that the pilot is not aware of they are occurring. Should the initial acceleration cue be too high or exist for too long, then the return acceleration becomes noticeable (false cues) and negative training may occur...as well a make the pilot nauseous.