FlyingWithoutLookingOut

Human balance and orientation depend on three systems working together: the eyes, the vestibular apparatus in the inner ear, and the pressure sensors distributed through muscles and joints. On the ground, with a visible horizon, they agree and the system works beautifully.

In cloud, at night, or over water with no discernible horizon, the eyes stop contributing. And the moment the eyes drop out, the other two systems produce confident, detailed, entirely wrong answers.

This is not a weakness that training removes. Instrument-rated airline captains are as susceptible to the underlying physiology as a student on their third lesson. What training provides is the discipline to disbelieve it.

The semicircular canals detect angular acceleration — the beginning of a rotation — by the movement of fluid against fine hairs. They do not detect sustained rotation. Roll into a turn and you feel it start; hold that turn at a constant rate and after roughly twenty seconds the fluid catches up with the canal and stops moving, and you feel perfectly level.

Now roll out. The fluid moves the other way, and you feel a turn in the opposite direction — from an aircraft that is now wings level. The instinct is to correct back into the original bank, which feels level and is not. This is called the leans, and it is the most commonly experienced form of spatial disorientation.

A related and far more dangerous version is the graveyard spiral. In a prolonged descending turn the sensation of turning fades. The aircraft loses altitude, the pilot pulls back to arrest the descent, and pulling back in a bank tightens the turn and increases the descent rate. Every correction the pilot makes on instinct makes the situation worse, and the aircraft spirals in while the pilot believes they are recovering.

Then there is the somatogravic illusion, which is purely about acceleration. The otolith organs cannot distinguish between the aircraft accelerating forward and the aircraft pitching nose-up — both press you into the seat back in the same way. A powerful acceleration after takeoff at night therefore feels exactly like an excessive climb, and the instinctive correction is to push the nose down toward a black sea.

The classic instrument panel arrangement — the six-pack — puts six gauges in two rows of three. What matters more than the layout is that they are fed by three completely different physical systems, and knowing which is which tells you what will fail together.

The airspeed indicator, the altimeter and the vertical speed indicator are pitot-static instruments. They measure air pressure. The airspeed indicator compares ram pressure at the pitot tube against ambient pressure at the static port; the altimeter reads static pressure alone and converts it to height; the vertical speed indicator watches static pressure change over time.

The attitude indicator, the heading indicator and the turn coordinator are gyroscopic. They depend on a spinning mass resisting change of orientation — driven by an engine-powered vacuum pump in older aircraft, electrically in most modern ones, and by solid-state sensors in glass cockpits.

The magnetic compass is its own system, and it is the only instrument on that panel that needs nothing from the aircraft at all. It also has significant errors in turns and accelerations, which is why it is the backup rather than the primary.

An instrument pilot does not read instruments one at a time. They run a continuous scan, and the structure of that scan is what keeps the aircraft under control.

The most widely taught structure is the selective radial scan: the attitude indicator is the hub, and the eye returns to it between every other instrument. Attitude, then altimeter, back to attitude, then heading, back to attitude, then airspeed, and so on. The attitude indicator is at the centre because it is the only instrument that shows what the aircraft is doing right now — every other instrument shows the consequence of what it was doing a moment ago.

The failure this defends against is fixation: locking onto one instrument, usually the one showing an error you are trying to correct, and losing the rest of the picture while you stare at it. An aircraft descending 200 feet below assigned altitude has a pilot looking at the altimeter, and a pilot looking at the altimeter is not looking at the heading, the airspeed or the attitude.

The other failure is omission — dropping an instrument from the scan entirely, usually the one that has been reading correctly for a long time. Instruments that have been reliable for an hour are exactly the ones that stop being checked.

Radio navigation began with ground stations transmitting a signal an aircraft could take a bearing from. The VOR, still in service, transmits on VHF between 108.0 and 117.95 megahertz and defines 360 radials outward from the station. A pilot tunes it, identifies it by its Morse code identifier, and can then fly directly to or from it along any chosen radial.

The instrument landing system is the precision approach that most airline landings in poor weather use. It provides two independent beams: a localizer giving lateral guidance to the extended runway centreline, and a glideslope giving vertical guidance down a path that is typically three degrees. Flown correctly, it brings an aircraft to a point a few hundred feet above the runway threshold with the runway ahead.

Satellite navigation has changed the picture more than anything since. GPS receivers augmented by ground-based correction can support approaches with vertical guidance to runways that have no ground equipment at all, which has brought instrument approaches to thousands of small airports that could never have justified an ILS installation.

What has not changed is the requirement to know where you are without it. Every navigation system has failure modes, and the pilot who can only navigate by following a magenta line has a single point of failure between them and the ground.