Navigating?

[Lump Jockey]

How do airliners navigate when it comes to great circle navigation? INS? GPS? Very interested in any answers on this!
LJ.
PS: How far out is the calculations when you get to your destination? Heard it's just a few feet! Do you have to re-set the system/INS/GPS?
TIA.

[Send Clowns]

INS, almost invariably. Some large jets still do not have GPS, for the rest there are still many areas of the world where it is illegal to use GPS as the primary means of navigation, notably the UK lest some of the private pilots forget they are still legally obliged to fix their position by visual reference under VFR or ground-based radio aids under IFR!

The INS position over the Atlantic can be out by a few miles (depends on the age and design of the kit to degree) but before landfall FMS (Flight Management System) position is updated by ground-based nav aids, reducing the errors to something very small. The error then depends on the radio aids used and the Kalman filter used to combine all the data being input to the FMS. The INS itself cannot be updated when the aircraft is moving, the last point of update is either on the stand or when take-off power is selected on the runway, depending on aircraft type.

Send Clowns
General Nav instructor,
BCFT

[Intruder]

I fly the 747-400 that is equipped with 2 GPS receivers and 3 INS units. All are tied into the FMS computer, and the FMS computes the track the airplane should follow.

In normal circumstances, the GPS position is used for primary navigation. The FMS keeps a second position as a backup, called the "IRS(3)" position, because it combines and weighs the position from each IRS (Inertial Reference System, sometimes called IMU or Inertial Measurement Unit), then updates it with DME, LOC, and/or VOR data when available. Note that the INS systems themselves are not updated -- the updating is within the FMS only. Also, GPS is not used for IRS(3) updating, to preclude excessive errors if both GPS systems go bad.

The IRS(3) position, as well as the position of each GPS and INS, can be read out from the FMS display when desired, and the individual plots can be put in the Nav Display.

GPS is the most accurate, and the Actual Navigation Precision (ANP) typically varies between 0.05 and 0.07 NM (300 and 420') -- less than 2 wingspans. Individual INS positions may be off as much as 3 or 4 miles after a 12-14 hour flight across the Pacific, and typically the errors run in the 0.1 to 0.3 NM per hour range (1 to 3 NM total). The FMS IRS(3) position is within about 0.2 NM when DME updating is available, and varies out to maybe half of the greatest individual IRS error enroute when the updates are stale.

[Lump Jockey]

Please, no sarky remarks now, as I'm obviously asking cos I don't know! But if they're out by so much, or so little as the case may be, how can the aircraft stay on course?:O

[Intruder]

The "courses" (airways) over the Pacific are 4 miles wide, and separated by 30 miles. It is usually only over open water, where there are no VOR/DMEs for updating, that errors more than a half mile are generated. Once in range of a VOR, DME, or LOC, the errors in the FMS are reduced again.

We stay so close to course that it's sometimes scary... Over a North Atlantic Track during "rush hour," a pilot may see 4 or 5 airplanes, all lined up perfectly in front of him, and several more on TCAS within 20 miles behind. Previously, before ring laser gyros and GPS, errors of a couple miles laterally were common, so the "big sky, little airplane" theory was credible. Now, the nav systems keep us so close to the magenta line that a controller error in altitude assignment, without an accompanying TCAS alert, is much more likely to have dire consequences.

[OBK!]

Modern IRS systems have a lot of redundancy in them, as some poeple on here may have pointed out. As well as accuracy. However, errors that do occur are most likely to be "bounded" errors. This means that the errors don't get bigger as the flight goes on. Instead, the errors will oscillate from there minimum value to there maximum throughout the flight.

The most "feared" unbounded error is Wander. I want get into that, I am not in too much of the know to explain it....yet!:O

I think the maximum error permitted is a deviation from real track of 3nm/hr. On arrival to the gate, the IRS position error is calculated. Say the FMS says you overall position error is 4nm, and the flight time was 12hrs then this is well within limits. I doubt it would be this much though. Maybe a few metres in most cases?