Wednesday, March 17, 2010

Something Can Be Done

The promise of Area Navigation (RNAV and GPS) was that it would be a simpler and more accurate way to navigate than older styles of navigation and to a great extent, that promise has been realized. GPS accuracy, especially when augmented with WAAS, is very good indeed. As for simplicity ... not so much. Waypoint navigation was a revolutionary concept when it was introduced, but it has been integrated with existing navigational paradigms and infrastructure in an evolutionary manner, not unlike the way an artist might sculpt clay or mold papier-mâché. This evolutionary approach has created some unfortunate and unforeseen complexity, but it doesn't need to be that way. Mom always said "Don't complain unless you can offer a solution or a suggestion," so here are my top five recommendations for simplifying the world of RNAV.

Wayward Waypoints

Many airports out there have a VOR located at airport and in those cases the VOR and the airport have the same name. Just as often the VOR may be some miles away from the airport, but both still have the same name. At the heart of every GPS receiver is a computer running software and software doesn't tend to handle ambiguity very well. That's why the GPS database encodes airports using a four-character ICAO identifier and VORs with a three-character identifier. The FAA's charting division could do us a big favor by using four-character ICAO airport identifiers on their chart products, but they don't. If they did, it would be crystal clear to student pilots and budding instrument pilots that KSAC refers to the surveyed center of the Sacramento Executive airport while SAC refers to the Sacramento VOR. 

Not in Kansas Anymore

The first step in GPS navigation is to enter the name of a VOR or NDB station on the ground, the name of an intersection of two VOR radials, an airport ID (which represents the surveyed center of the airport), a charted VFR reporting point, a Computer Navigation Fix defined by FAA chart designers, or even a user waypoint that you've created. The AIM refers to this type of navigation as to-to, not to be confused with Toto, the little black terrier in the Wizard of Oz. GPS receivers only navigate to one waypoint at a time, also known as the current waypoint.

GPS makes it simple to navigate to a waypoint and most receivers provide a moving map display, which is a score for simplicity and safety. The bad news is that unless you're lucky enough to have a keyboard as part of your GPS receiver, entering a waypoint requires a precise and often convoluted sequence of knob-turning and button-pushing. A bad user interface makes it all too easy to misspell the name of the waypoint: Get just one letter wrong and instead of navigating to a VOR that is 20 miles away, you may be headed to Tierra del Fuego by mistake!

The engineers that designed GPS receiver user interfaces didn't set out to create difficult-to-use products, but the fact is they did. Whether it was the desire to save a few bucks by having fewer knobs and buttons or simply a race to get a product to market, it's clear that mistakes were made. Now the users of these products have to live with the mistakes and to quote Warren Zevon, "… it ain't that pretty at all." Bad UI design is the Achilles heel of GPS and many of us pilots have become so acclimated to these convoluted interfaces that we have lost sight of just how whacked this situation is.

Near the top of my "need to fix" list is Garmin's Small-Knob/Big-Knob interface. You press the small knob to enter "cursor mode" so you can edit or enter the name of a waypoint in a flight plan. You turn the small knob to start the process of entering letters and then the small knob changes function. Whoa there! A knob whose function changes depending on an interface context that is mostly invisible to the user? This needs to be fixed and one simple way would be a separate button dedicated to starting and ending the waypoint editing mode.

Having a separate button for edit mode would also fix the problem that countless new Garmin users run into: Pressing the small knob to exit cursor mode and accept whatever changes they have made. Having watched hundreds of pilots make this mistake thousands of times it's clear that a common intuitive belief is that if you press one button or key to enter a mode, pressing the same button or key should exit that mode. In the Garmin world, this simply exits the editing mode and, here's the amazing part, destroys whatever changes you made without asking you to confirm that's what you want to do. This is B-A-D.

Missing Pieces on the Missed Approach

When flying an instrument approach, most GPS receiver are designed to suspend the automatic sequencing of waypoints when you reach the missed approach point. Think about this for a moment: You're close to the ground with reduced obstruction clearance at a high-workload moment. You're either going to see the runway and land or you won't see anything and you'll fly the missed approach. Is this really the time to make a pilot divert their attention from controlling the aircraft to push the OBS button or SUSP soft-key? I don't think so and apparently neither did the designers of the GNS 480 (nee CNX 80), which will automatically sequence to the missed approach segment. If you see the runway environment and decide to land, you just ignore the GPS. If you don't see the runway environment or loose sight of the runway while circling, use the GPS to start navigating on the missed approach. Too bad the GNS 480 is out of production and the GPS units that are in production don't exhibit this behavior. A defense I've often heard is that the TSO specifies that pilot action is required to initiate the missed approach and if this is true, the TSO should be changed.

When flying a non-RNAV approach, many GPS receivers automatically switch the navigation source from GPS to the VOR or localizer receiver. That's great, but if you need to fly the missed approach and you want to use the GPS to do so you must divert your attention and manually select GPS as the navigation source. I mean really! If it's okay to automatically switch navigation source out of GPS, why not back into GPS mode?

Four Card Minima

There's a new game for RNAV approaches that all pilots must play and it's called "Guess the approach minima." It goes something like this. When you brief an RNAV approach, you may see up to four sets of minima listed: LNAV, LNAV/VNAV, LPV and circling. The issue is you may not know which minima your WAAS GPS receiver can offer (based on current signal integrity) until a few miles before the final approach fix. This has to do with the design of WAAS GPS receivers' final signal integrity check and I honestly can't think of a good way around this shortcoming: You just have to brief multiple approach minima and choose the correct minima based on the approach sensitivity your WAAS GPS receiver displays.

Where improvement could be made would be to ensure that the approach sensitivity displayed on the GPS receiver exactly corresponds to the approach minima shown on the chart. If your receiver arms with  LPV or LP sensitivity, you're okay because your WAAS GPS receiver should display LPV or LP. If the receiver arms with LNAV sensitivity, you may see LNAV or LNAV+V. If it arms with LNAV/VNAV you'll probably see LNAV/V. Notice the subtle, similar appearance of LNAV+V and L/VNAV? This is too subtle and is B-A-D. And the minima shown on the charts should exactly match the minima displayed on the GPS receiver, period, end of discussion.

Procedure Turn or No?

The introduction of the Terminal Arrival Area (or TAA) was meant to simplify pilot/controller interaction when executing an RNAV approach. And it would be simpler, were in not for the fact that not all RNAV approach charts follow the same conventions with regard to the depiction of a hold-in-lieu-of procedure turn (or HILO). In particular, some RNAV approaches have a standard Minimum Safe Altitude circle while others display minimum safe altitudes in sectors on the plan view of the chart. The subtle problem is that MSA sectors will usually tell the pilot that the procedure turn is not authorized when you're headed straight-in to an Initial Approach Fix where a HILO is depicted, while approaches with the MSA circle do not.

If you were approaching from the Southeast and were told "when able, proceed direct HERMIT, cleared RNAV 34 approach" you need to read the fine print on the MSA sector shown on the plan view to know when you could descend and to know that the HILO is not authorized.

If you were told "when able proceed direct CADAB, cleared RNAV 29 approach," you need to know that the hold-in-lieu-of procedure turn is required unless the the controller remembers to say "… cleared straight-in RNAV 29 approach."

The FAA charting division needs to come up with a consistent way of depicting MSA and clearly denoting when a procedure turn is required and when it isn't. Until then, pilots should ask the controller when they see a HILO and they aren't sure whether or not they are expected to fly the procedure turn.

More Fond Wishes

So that's my wish list of the top five features and enhancements I'd like to see for the world of RNAV. You may have your own list of desired features, too. For now, we can only hope that the people in a position to fix these issues are listening.


Millz said...

Great post - in particular the issue with the Garmin losing your input when you exit cursor mode has happened to me far too many times. The Garmin folks seem to think more like software engineers than users in most of their UI.

Jeff said...

Another thing that might simplify GPS device use would be to allow programming of waypoints outside the airplane. E.g. determine the waypoints on your desktop computer, export to an SD card or something, and then insert the SD card into aircraft GPS. One could export information that way from the aircraft, too... actual flight paths, times, distances, etc.

ddf said...

This is a GREAT post John, thanks. I am one of the idiots pushing and turning, only to find that my flight plan is all hosed up. This is especially frustrating while circling in the missed approach holding pattern preparing for an approach into an alternate. Having gone through the extensive workload just to get into holding, its insult to injury to find that one wrong push eliminates all of the work done to setup the new procedure. ...and all this time I thought it was just me.

Marshall said...

The lack of auto-sequencing for missed approach is my biggest fear in instrument flying. If I had to go missed in actual instrument my plan is to guess at a heading until I'm satisfied I've got some altitude before I start poking at the OBS button. (I fly a G1000)

Anonymous said...

Another gotcha is in determining when to enter the airport identifier prefix ('K' for US airports, 'Px' for Alaska and Hawaii airports).

The Airport identifiers that only contains alphabetic characters are entered in the ICAO format with the prefix, while airport identifiers that contain alphanumeric characters are entered as shown in the FAA charts. IMHO, this also adds to the confusion and should be addressed.

Ron said...

I guess I'm playing devil's advocate here again, but I've never been bothered by the automatic suspension of waypoint sequencing at the missed approach point.

I agree that when the airplane is slow and low to the ground, one should not be worrying about button pushing. So why do it?

What I teach students when briefing an instrument approach procedure is to memorize the first part of it. They always start the same way: climb. So that's easy. Then all you have to know is how high are you going to go before starting a turn (if any).

In my experience, that gives the pilot plenty of time from the MAP to apply full power, pitch for climb, raise flaps and/or gear, and trim the plane.

That last part is very important. TRIM THE PLANE. People run into problems because they are low, slow, and not trimmed. Once the airplane is trimmed for climb, you're probably at least a few hundred feet above minimums. THEN you can push the OBS key to un-suspend waypoint sequencing.

I actually LIKE the fact that the OBS key has to be pressed, because it helps me teach students what's really important at a critical phase of flight. All students seem to initially want to press the OBS key before anything else, to the virtually uniform detriment of their flying. Even if the GPS auto-sequenced to the next waypoint (as in the case of the ex-Apollo 480 receiver), that desire to return one's primary attention to the GPS would still be there right at the MAP. I want to break students of that habit. You can fly the first few critical seconds of a missed approach procedure without the GPS, and it's far safer to Just Fly The Plane at that point.

Regarding SD card input of waypoints prior to reaching the aircraft, that's a great idea. I think the FAA and OEMs are probably not keen on the idea because they are wary of bad data somehow corrupting the GPS receiver (I've seen that happen in-flight due to a bad database update, and it's not pretty). In an era of viruses and other software issues, they want the GPS system to be as "closed" as possible.

Gatorpilot said...

Very interesting. I was thinking 'but... but... ' in the back of my mind because I'm so accustomed to some of these idiosyncracies now that to do it another way seems out of place, but the truth is you're probably right on the mark for most of these.

1) "Wayward Waypoints." Yep, although it is incumbent upon us to understand it, it's still confusing to newer pilots.

2) Garminology. Probably also a good suggestion. I taught behind the 430 and 530 for years, and now teach the G1000 among others, so to me this is now automatic. However, I too have seen many students have trouble with this.

3) Automatic switch to missed approach profile. I will disagree here because there is a world beyond little airplanes and for many reasons I would not want my equipment to auto-sequence to the missed approach segment just because I cross the MAP. When I'm flying bizjets, I hit the GA button upon reaching the MAP which commands many things to happen -- the flight director command bars pitch up to 12 degrees, I get a GA Power bug on my N1 indicator, and the FMS sequences to the missed. I want all of these things to happen at one time as one configuration change, not piecemeal, and there's also the logic of determining how the unit would auto-sequence at DA in the case of an ILS approach. Further, there are times when one might be flying an alternate missed (think single engine, runway analysis, climb gradients). For many reasons, I think pressing a single button is not unreasonable, and the pilot should initiate this action. One notion I would support is that perhaps this button should be placed on the yoke or power levers/throttles, but I've never had much difficulty in getting my students to simply press the button wherever it happens to be located.

For that matter, consider some of the advanced factory and aftermarket panels that are becoming increasingly available in smaller piston airplanes. Flight directors will eventually find their way into most if not all production IFR-capable aircraft, and we'll probably see similar combined logic in the GA mode. You don't want this sequencing automatically, especially in that type of aircraft, with lower-time pilots.

4) Multiple minima. Can be confusing, definitely. I think the confusion lies more in the terminology than in how the WAAS receiver measures/displays RAIM, but I agree with the thrust of your comment.

5)TAAs. The example you chose (O28) is interesting because Jepps are MUCH more confusing when it comes to multiple altitudes and requirements for NoPT on the transition. The Jepp example would show a list format and only denote "NoPT" beside the FIRST altitude, but it's assumed the pilot understands it means NoPT for ALL altitudes. Very confusing indeed.

It's not often the government does a better job than the private sector, but they do in this case.

I'm not sure how I feel about this one -- probably neutral. I think the segments, IF/IAFs, and TAAs are pretty clearly labeled "NoPT" and the AIM is very specific on this subject as well. I'd probably leave this one alone as we'd likely only make it more confusing by changing it now.

John Ewing said...


Climbing on the missed, sure. But a button push is NOT needed to make someone realize which phases of flight are critical. And besides, it's not always just a single button push.

Reaching the MAP on a non-RNAV approach when flying a G1000 aircraft requires - count 'em - three button pushes - Once on the SUSP softkey, twice on the CDI key. This is hokey! It's kooky! And it doesn't have to be that way. And don't get me started on the subject of the crazy finger dance required on the G1000 if you need to start the missed approach early.

Not allowing user-defined waypoints to be programmed on another device and then imported has not prevented DB corruption issues, as witnessed by the Bendix-King DB errors that had to be corrected just a few weeks ago. Data/memory protection concepts have been around for, oh I don't know, at least four decades. The idea that user data couldn't be kept separate from "official" data assumes facts not in evidence.


A go-around feature like you describe is a design I would like to see in light, GA aircraft, but it ain't there in most. With two crew members, a few more button pushes aren't such a big deal. Single-pilot IFR with a guy or gal who is fortunate to be able to fly once or twice a month? That's a safety issue and it doesn't have to be that way ...

If nothing else, these two comments illustrate how pilots who become used to "the way it is" can lose objectivity about just how bad "the way it is" really is.