Mentor Matters: Glass cockpits and ILS ‘auto-switching’

The ‘101st time’ can be a bear for the unwary

October 1, 2013

In this example, the pilot remains at his clearance altitude of 3,000 feet, hoping that he’ll capture the ILS glideslope from below—somewhere between RUNTS and HOBEL. But because the airplane is above the glideslope intercept altitude for both fixes, automatic autopilot capture of the glideslope won’t happen. Indeed, the fact that the aircraft is above the glideslope will be hidden from the pilot until passing RUNTS. If slow on the draw, the pilot ends up high, fast, and unstabilized on the final approach course. In this case, the best option would be to execute a missed approach rather than try to rescue the approach.

The integrated avionics suites installed in production light jets are true marvels of automation and workload reduction. Complex procedures that in older jets would require multiple pilot actions are completed by modern flight decks with a single button push: e.g. switching guidance from an ILS to the missed approach and starting the sequencing on the missed, or making multiple stepdowns on a non-precision approach. Likewise, the nearly everyday action of transitioning from en route to ILS guidance is largely automated; the flight deck will switch at the appropriate time from “magenta needles,” where GPS information is driving the HSI, to “green needles,” with the ground-based ILS signal providing information to the HSI.

This automatic transition normally happens so seamlessly, and without pilot intervention, that pilots can quickly become accustomed to not being aware of it. Like the act of breathing, it happens in such a behind-the-scenes manner that the fundamentals of how it occurs aren’t considered. Certain approaches, however, are built in a manner that can confound the flight deck, and hence the pilot, if careful forethought and avionics monitoring does not occur.

The ILS Runway 31C into Chicago Midway (MDW) is a prime example. Approaching Chicago from the east, a pilot often will fly an arrival that brings the aircraft very close to the final approach course of the ILS 31C. Depending on traffic load, the pilot may be vectored onto the approach, or be cleared to fly Direct to one of the waypoints outside the final approach fix (FAF). It is this latter clearance that can set the pilot up for the trap in question, and leave a pilot scrambling to descend on the glideslope.

The dynamic at play here is the method used by the flight deck to know when to transition from magenta to green needles. Most modern flight decks, regardless of manufacturer, do so when the FAF becomes the “to” waypoint. Stated otherwise, if flying a chain of waypoints before the FAF, the system will not auto-switch to green needles until passing the fix immediately before the FAF.

Imagine a jet is approaching MDW from the east, and receives the following clearance: “Proceed direct GLEAM, cross GLEAM at 3,000 feet, cleared for the ILS 31C.” This type of clearance is very common when flying an ILS, and a pilot is normally able to simultaneously: proceed direct to the assigned fix (via GPS mode), descend to the assigned altitude (typically in vertical speed mode), and arm the approach mode of the autopilot. Crossing the fix prior to the FAF, the system will switch from magenta to green needles, and as the AP approach mode was already armed, capture the glideslope (GS) when it is intercepted.

The unusual geometry of the MDW ILS 31C makes this method unworkable, however, without further pilot action. Crossing GLEAM at 3,000 feet, the system will still be in magenta needles, and the pilot is able, but not required, to descend to 2,500 feet. Many pilots will not step down the extra 500 feet in vertical speed mode, but would rather elect to simply wait until the GS is captured to descend. For the vast majority of ILS approaches this technique will work well, in that it removes two power changes from the approach, simplifying airspeed management.

Doing some quick math, it’s apparent why waiting for GS capture won’t work here. Given that the GS crosses HOBEL, the FAF, at 1,700 feet, and that a normal three-degree GS descends approximately 300 feet per nautical mile, we can see that as RUNTS is 2.5 miles from HOBEL, the GS will intercept RUNTS at 1,700 feet + (2.5 times 300 feet), or 2,450 feet. If the pilot plans to remain at 3,000 feet until capturing the GS, he’ll be in for a rude surprise; as the system can’t capture GS until in green needles, and as it won’t be in green needles until passing RUNTS, by the time the system is ready to capture GS, the aircraft will be more than 500 feet too high to do so. Modern autopilots will not automatically capture a GS from above, so the pilot will be forced to take manual action to force the aircraft to descend rapidly and “catch” the GS. If the pilot is aware enough to immediately understand what has happened, an extra several hundred feet per minute of vertical speed should be enough to capture the GS by the FAF, but with any delay on the pilot’s part the aircraft will end up unstabilized and high until well inside the FAF.

As unstabilized approaches are a causal factor in many runway overrun accidents, ending up high (and as a result, often fast) is to be avoided at all costs. Should pilots find themselves in this situation, the proper action is to perform a go-around, and reattempt the ILS approach. Unfortunately, studies have shown that less than 10 percent of unstabilized approaches end in a go-around; pilots more often suffer from tunnel vision, fixating on “saving” the approach.

Given the above, a prudent pilot should remember that there are outliers for nearly every type of operation in modern aircraft. We can do something a certain way one hundred times in a row, and if we don’t keep in mind the “whys” of what is happening, the 101st event could have an unusual set of variables present, which will completely throw our habitual actions into disarray. Briefing this approach, a careful pilot would catch that the distance from the FAF to the fix immediately preceding it is much smaller than typical.

Combine this with good knowledge of why and when the avionics will switch navigation source, and the pilot will be able to anticipate trouble ahead of time. Knowing that the automatic switch to green needles will occur too late on the ILS 31C, the pilot can manually make the switch well before RUNTS. Doing so, the AP will be in a position where it can capture the GS from underneath, and a stabilized approach can be performed.