I read that an ILS is a precision approach system. It provides a final approach with precision in both horizontal and vertical guidance to the runways. But I don't know how precise it is because there were a number of accidents related to landing operations.

  • $\begingroup$ I would assume that it is extremely accurate, as autolanding things are able to land it pretty much perfectly every time $\endgroup$ May 17, 2014 at 23:32

3 Answers 3


The ILS is precise enough to allow landings in essentially zero visibility, including those performed by autoland.

The ground based bits

The vertical range is 1.4° (0.7° above glideslope, 0.7° below) and the horizontal varies, but is a maximum of 6° wide. What this means in terms of how many feet or meters off the centerline an aircraft is varies depending on where the aircraft is on the approach, but it's generally correct to say that the closer an aircraft gets to the ILS antennae, the more precise the system is.

The specifications for a localizer antenna require that it be no less precise than ±10.5m - this is for a Cat I approach, and it gets correspondingly more precise as the category increases.

There's always the risk of interference with the transmitters; as noted in the comments, accidents have been caused by airport vehicles parked in the ILS critical area, interfering with the signals, and as the equipment ages it can require recalibration.

The airborne bits

It's possible for receivers to have issues that could affect precision; antennas can get loose connections or become covered in ice, while pilots might interpret indications incorrectly, misread reverse sensing, or intercept a 'false glideslope' from above.

Regarding accidents

You're correct - there are many accidents related to landing operations. That said, you may be including those which weren't flying an ILS approach - it's hard to say since you aren't citing anything. Still, it's possible that there's confusion between navigation precision and safety of flight.

Many (not all) of landing accidents involving instrument approaches do center around human error rather than equipment failure; pilots tuning in the wrong frequency or incorrectly setting avionics and instruments (reversing the inbound course on an HSI causes reverse sensing, that sort of thing), or people on the ground screwing up.

  • 2
    $\begingroup$ There's also the issue with false lobes. $\endgroup$
    – falstro
    Feb 12, 2014 at 17:09
  • 2
    $\begingroup$ @roe true, but I'm not sure if that falls into the 'precision' category ... technically, the false lobes are just as precise as the correct ones. They're just... wrong. Inaccurate yet precise, if you will. $\endgroup$
    – egid
    Feb 12, 2014 at 17:10
  • 1
    $\begingroup$ Precisely inaccurate? :) I don't know if there are any accidents attributed to false lobes, but it could potentially, with all its precision, guide you down in reverse sensing steep glide slope, causing confusion resulting in loss of control for example. $\endgroup$
    – falstro
    Feb 12, 2014 at 17:12
  • $\begingroup$ Also, although the question says "precision", I think it is fair to say the answer "it's precise, just not pointing you where you want to go" is splitting a few hairs... :) $\endgroup$
    – falstro
    Feb 12, 2014 at 17:17
  • 7
    $\begingroup$ The major issue with ILS accuracy (or for that matter precision) is there's no way to measure it from the aircraft: The ILS gives you two needles (Fly Up / Fly Down & Fly Left / Fly Right), but it has no way of telling you if the signal has been perturbed by say a truck parked next to the ILS antennas (which has been known to happen). Signal attenuation from that sort of thing mainly affects accuracy, but could conceivably affect precision as well. $\endgroup$
    – voretaq7
    Feb 12, 2014 at 17:40

ILS is can be very precise, but there are many factors that may compromise the precision. When all these factors are avoided, automatic landing with no visibility is reliably possible.

The ILS installations are classified in three categories:

  • CAT I: Category 1 allows for decision height (DH) no less than 200 ft (height, so above ground level).

    ILS creates the signal by interference of different signals emitted by offset antennas. If one of them fails, but not the other, the signal may appear as perfectly centred to the instrument on board of the approaching aircraft with no way to tell that it does not actually work. Therefore pilots should cross-reference their position using other sources (DME, VOR, ADF, marker beacons).

    There is a video about this kind of failure (describes an actual incident where the glide slope stopped working without indication, the crew noticed by cross-checking against DME and altitude, plus description of the failure), but I can't find it now.

  • CAT II: Category 2 allows for DH not less than 100 ft and allows using auto-land until touch-down. It needs to have far field monitoring that can detect incorrect or incomplete signal and the transmission system has to be doubled. It also needs to be somewhat more precisely aligned than CAT I.

  • CAT III: Category 3 allows auto-land including roll-out. It has subcategories IIIa that allows DH not less than 50 ft, IIIb that allows DH down to 0 ft (so effectively no DH) and visibility 75 m (otherwise the plane could not taxi off the runway as there is no guidance for taxiing; when there is, it will be called CAT IIIc). It has stricter limits on precision and monitoring than the CAT II.

Independent of the category, direct line of sight is needed. The signal is not affected by fog and precipitation, but is distorted by any obstacles. Aircraft or even small vehicle crossing the runway or taxiing in front of the transmitter will significantly distort the signal.

That's why each ILS CAT II and CAT III installation has ILS protected areas and to actually use auto-landing, special procedures need to be in effect that keep these areas clear. When they are not, incidents like Singapore B773 at Munich on Nov 3rd 2011, runway excursion happen.

Similarly attempting to auto-land on CAT I lead to accidents like Air India A320 at Jaipur on Jan 5th 2014, runway excursion on emergency landing.


Quoting egid:

"The vertical range is 1.4° (0.7° above glideslope, 0.7° below) and the horizontal varies, but is a maximum of 6° wide. What this means in terms of how many feet or meters off the centerline an aircraft is varies depending on where the aircraft is on the approach, but it's generally correct to say that the closer an aircraft gets to the ILS antennae, the more precise the system is."

I know of at least one reader who has a problem with this statement, so I'm only adding these comments for clarification. EGID is correct. Think of the localizer as a funnel ... the "hopper" end of the funnel catches more liquid when it is bigger and fans out wider ... but it ultimately leads to the same narrow downspout. If the purpose of this funnel is to catch and direct more incoming liquid (aircraft) then it's doing no one favors to have too narrow of a hopper. All funnels are very precise on the small end. You want an ILS to be fairly forgiving for the reason that you may be intercepting it while blind and in the soup.

  • $\begingroup$ The localizer is placed at the far end of the runway (if it was at the approach end, it would become very erratic close to touchdown, and we do not want that), hence it cannot diverge more than to cover the runway at the other end, or you'd be coming down on a taxiway, or park it directly in a hangar. This doesn't really matter if your needle is both accurate and precise and you're able to keep it centered, I just wanted to point out that the touchdown point is not at the 'hopper' end of the funnel of the localizer (glide slope is another matter though), it's actually a couple of miles away. $\endgroup$
    – falstro
    Feb 13, 2014 at 7:36
  • $\begingroup$ @roe: The main reason why localizer is at the far end is that it still needs to indicate correctly while you are over and on the runway. $\endgroup$
    – Jan Hudec
    Feb 13, 2014 at 16:14
  • $\begingroup$ @JanHudec; yes, that's what my first sentence said (check the parenthesis) :) The rest still holds, the touchdown point is nowhere near the end of the funnel, and if the precision isn't good enough, you'll end up off the runway, so it isn't all "the same narrow downspout" $\endgroup$
    – falstro
    Feb 13, 2014 at 16:16
  • $\begingroup$ @roe: I checked the parenthesis. But I am saying it's not it would become erratic, it would flat out not be available at all. $\endgroup$
    – Jan Hudec
    Feb 14, 2014 at 9:38
  • 1
    $\begingroup$ @JanHudec; You're right, it would be unavailable for a short while over the antennas before capturing on the back course (being reverse sensing at that point). Don't know for how long it would be silent, but it'd probably be available again before reaching the touchdown zone, no? Either way, it's a bad idea, and completely beside the point I was trying to make. :) $\endgroup$
    – falstro
    Feb 14, 2014 at 10:23

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .