Thursday, September 09, 2010

GPS Transition, Part III: Conservation of Complexity

Some folk seem to think that if GPS navigation is simple, then something must be done to make it more ... difficult.  Let's call this the Conservation of Complexity Principal: An area of aviation reaches a level of complexity causing pilots, instructors, examiners, inspectors, and even engineers and designers to feel compelled to maintain or even increase that level of complexity. Simplicity should be a primary goal in systems design and it is unfortunate when we mistake increased complexity for progress. Newer glass panel aircraft have a ton of features, some crucial, some cool, and some seldom-used “gee whiz” features. RNAV procedures and equipment are also cool, they provide much flexibility and increased accuracy, but they are mired in a confusing swamp of details. Here are some thoughts on embracing the essentials while eschewing the geeky.

Can You Top This?

Needless complexity comes into the picture when equipment and procedure designers, probably under time pressure, jump to a quick solution. Some hardware manufacturers, hoping to establish market differentiation, end up designing too many features into their products. Some pilots, instructors and examiners embrace this geeky stuff without any clear evidence that knowing lots of minutiae in any way enhances learning, proficiency, or safety. Requiring pilots to understand and recite facts about the GPS satellite constellation or to describe the Line-Replacable Units that make up the G1000 and how they communicate with one another has few clear advantages, other than making it easy to test a candidate’s knowledge of the nitty-gritty.

It may be interesting to know the model designations of each G1000 component and the network media used to connect them, but what does that have to do with effectively managing the technology in newer aircraft while flying single-pilot? At it's worst, teaching detailed systems knowledge may encourage pilots to try to figure out why something has failed while in flight, resulting in a distraction that can adversely affect the safety of flight. If a GPS or one of its components fails during flight, the pilot needs to recognize the failure and understand the extent to which their ability to navigate has been affected. Then they can get the aircraft safely on the ground and have an avionics tech diagnose and fix the problem. Acquiring and maintaining complex, systems-level knowledge of the G1000 is not needed unless you plan on repairing the hardware yourself.

Papier-mâché Unicorn

RNAV procedures and policies have grown up over time with the added challenge that they be made to exist with existing VOR/NDB procedures and policies. Add a little here, a little there and after a decade you stand back to look at what’s been created. Voila! It’s a unicorn! In defense of the FAA, a tremendous number of new RNAV procedures have been created and the majority of them for other than Part 139 airports (i.e. GA airports). Nevertheless, mistakes were made.

Charting complexities add tiny bits of workload in a single-pilot environment. Approaches have titles like RNAV (GPS) RWY 22, but then there are obscure rules that say any item contained in parenthesis is to be omitted. There are variations and inconsistencies in terminal arrival areas, the depiction of minimum safe altitudes, required equipment, important details buried in notes, and even special climb performance on the missed approach. Trying to keep all of this straight while flying single-pilot may seem like death by a thousand cuts and can contribute to creeping task saturation. Here’s an approach for an airport in my area that was published and then quickly NOTAM’d as not authorized. Hmm ...

Need to Know

GPS units like the G1000 offer a plethora features, leading many pilots to mistakenly believe they need to understand all of the bells and whistles. Many pilots see this morass of details and just turn away. So here’s some heresy for you: You don't need to understand every single feature. It is possible to get along just fine without knowing how to get your GPS to add an along-track-offset or display a parallel track. Once you understand the basic concepts of waypoint navigation, GPS flight plans, automatic waypoint sequencing, turn anticipation, and how to load procedures, you can hone your skills using the concise list of tasks that every pilot should be able to accomplish with their GPS under instrument flight rules, found in the AIM (1-1-19(p)).

  1. Perform a RAIM prediction function
  2. Insert a DP into the flight plan
  3. Program the destination airport
  4. Fly an overlay approaches (especially procedure turns and arcs)
  5. Change to another approach after selecting an approach
  6. Fly "direct" missed approaches
  7. Fly "routed" missed approaches
  8. Enter, fly, and exit holding patterns
  9. Program and fly a "route" from a holding pattern
  10. Program and fly an approach with radar vectors to the intermediate segment
  11. Handle a RAIM failure both before and after the FAWP
  12. Program a radial and distance from a VOR (often used in departure instructions)

Devil in the Details

At the heart of every GPS receiver is a computer running software and as the late John Ciardi said, "Computers are high-speed morons." GPS database designers encode airports using a four-character ICAO identifier and VORs with a three-character identifier. In some cases, the VOR is located on an airport with the same name and identifier, often not far from the surveyed center of the airport. The FAA's charting division needs to clean up their act and use four-character ICAO airport identifiers on their chart products. This one simple change would make everyone’s life easier, but as of this writing the Aeronav folks haven't gotten around to it.

If you fly RNAV approaches with an IFR-certified WAAS GPS, you’ll need to be prepared to fly up to one of three different minima on an approach - LPV, LNAV/VNAV, or LNAV. You also need to correlate the course sensitivity a GPS displays during an RNAV approach to the correct minima shown on the approach chart (LPV = LPV and LNAV/V = LNAV/VNAV). If your GPS is not WAAS capable, you’ll only get LNAV minima and your life is simpler.

This brings up a question that many renter pilots have a hard time answering: How do you know if the plane you just rented has a WAAS GPS or a non-WAAS GPS? There’s nothing on the face of the GPS unit that tells you this, you have to consult the Approved Flight Manual Supplement.

Another way is to bring up the satellite status display and if you see a “D” in one or more of the satellite signal strength histogram bars, the unit is a WAAS GPS. What does the “D” stand for, you ask? “Differential GPS,” another subtle and annoying mental connection you must make. On G1000 models you also see the term SBAS (Satellite-Based Augmentation System) instead of WAAS.

Non-WAAS GPS units detect navigational errors with a process called RAIM (Receiver Autonomous Integrity Monitoring), which is analogous to seeing a flag appear on your VOR course deviation indicator or HSI to indicate the station is unusable. WAAS GPS units use a different error detection process called FDE (Fault Detection and Exclusion), another good reason for knowing exactly what kind of unit resides in the instrument panel of the aircraft you just rented or may be thinking of buying.

Updates to GPS databases are made available every 28 days, database subscriptions are not cheap, and downloading and applying the updates can be time-consuming. These are the costs of doing GPS business and if you fly at night, over remote areas, or in the clouds, you don't want an out-of-date database. Many feel that the costs of database updates has gotten out of control, especially when an aircraft owner must have multiple database subscriptions for the same model of GPS unit installed in the one aircraft. Glass panel aircraft may have numerous other databases, often with differing expiration dates. When the tension between being safe and legal collides with saving time and money, databases don’t get updated and renter pilots need to be prepared. The plane you just rented may have lots of glass in its panel, but if the GPS aeronautical database isn’t updated the plane goes from “/G” to “/U”

Wither ERAM?

Last, but not least, is the FAA’s continuing struggle to provide IFR clearances that take advantage of the capability of the most basic GPS receiver. One of the problems has to do with FAA flight plan formats because there is only so much information about an aircraft’s type of RNAV equipment that can be encoded in a single letter, like “/G.” ICAO flight plans provide much needed detail on just what level of RNAV capability an aircraft does or does not have. ICAO flight plans are now required for aircraft that desire RNAV departures. Even if you don’t file an ICAO flight plan, you can request and get an amended clearance to proceed direct to a waypoint or VOR once you’re airborne, but that means reprogramming a GPS flight plan in flight. Not just occasionally, but every time.

At the crux of this problem is the FAA’s computerized system for generating clearances that hasn’t changed in years. The new ERAM (En Route Automation Modernization) system was supposed to change all that, but it doesn’t seem to be working out very well at the moment. I’ve tried filing all sort of routings out of KOAK using ICAO flight plans with various routing strategies, but I always get the same routings. Having a technically advanced aircraft with all sorts of accurate, flexible navigational capabilities doesn’t count for much if the ATC computers are not prepared to play ball.

The best bet still seems to be using your local knowledge of what ATC usually assigns or use one of the many briefing services (, ForeFlight, that can tell you what routing has recently assigned to your destination. Use that routing on your flight plan and once airborne, start asking and negotiating a more direct routing with the controller. Kasparov may have been beaten by a chess-playing computer, but it seems ATC professionals are still more flexible and creative in the aviation chess game.

Hits Just Keep Coming

Now that you understand the basics of GPS navigation, the features that were supposed to make your piloting tasks easier, you should be better equipped to learn the complexities of your particular model of GPS so you can bend it to your will. Stick with the basics, practice those tasks listed in the AIM, and you should have fewer moments of GPS confusion. You might even find that if you hold you nose, bite your tongue, and cross your fingers you’ll find that GPS navigation is not really all that bad. Eventually one hopes that the FAA and the equipment manufacturers will see the error of their ways, fix the screwed up user interfaces and develop strategies to smooth out the significant RNAV wrinkles (or at least keep more wrinkles from appearing). Until they do, take heart in what Dos Passos observed: There are things that could be more, but are content to be less.


David said...

I'm having trouble understanding why you used the Napa approach plate in your example. Do you mean to imply that the MAP caveat is misplaced? 250 ft/nm isn't terribly difficult unless you either have a really fast ground speed or if you're totally iced up ...

John Ewing said...

There are two reasons I used the APC RNAV (GPS) Z 36L as an example.

First is the approach chart's raison d'être - a Z approach simply because the missed approach requires a climb rate greater than the standard 200'/nm.

The other is the fact that no sooner was the approach published, it was NOTAM'd as not authorized. What's up with that?

The first time I was asked why there were two RNAV approaches to RWY 36L at Napa (APC RNAV (GPS) Y RWY 36L and APC RNAV (GPS) Z RWY 36L), I had to study both charts for a couple of minutes before ferreting out the reason. In the Jepp version of the Z chart, the climb rate is buried in the notes section and easy to miss.

That's kooky, if you ask me.