Syn Vision For Approaches

  • E-Mail this Article
  • View Printable Article
  • Text size:

    • A
    • A
    • A

The other evening I was returning from a flight to the east coast of Florida in an aircraft Sarasota Avionics kindly loaned me for the day. I landed just past dusk when, rare for Florida, the temperature and dewpoint came together and as I was tying down, the fog was forming. We occasionally get marine layers off the Gulf, too, but if that happens four times a year, I’d be surprised.

The best the approaches into Venice can do is 360-foot MDAs and a mile of required visibility. It’s unlikely they’d be of much help in dense fog. For the past three days by early evening, the ASOS had been reporting 200 feet and a ¼ mile, if not indefinite. I’m glad I made it in before it got worse. “Well,” my friend Dan mentioned half seriously, “you’ve always got the synthetic vision.”

True enough. The airplane has a pair of Aspen EFD1000s, with synthetic vision. My initial reaction was that I’d never use syn vision to land or carry on below MDA, but after I thought about that, I asked myself why. The kneejerk answer is that FAR 91.175 is so baked into to me and as an instrument instructor I’ve taught its thou-shalt-not religion for so long that I don’t know anything else. In case you’re rusty on your FARs, 91.175 requires the runway or its environment to be in sight before descending below MDA or DH.

There are technical reasons the regulation is worded the way it is and it isn’t based on some bureaucratic capriciousness. Instrument approaches are designed to be and are 100 percent guaranteed safe if you fly them exactly the way they’re charted and if the equipment you’re flying is functioning correctly. The obstacle clearance is carefully considered with margins for system and human error. The FAA routinely flight checks procedures looking for anomalies. But once you depart the black lines, you’re on your own.

To find a runway out of the clag, however, you necessarily have to fly down a funnel to the minimum obstacle clearance at which point you’re supposed to be able to see the runway visually and land safely. For a standard CAT I ILS, that’s only 200 feet and some GPS approaches offer MDAs nearly as low as that. Two hundred feet isn’t much, which is why instrument pilots who hope to survive to a long and rewarding career best not make a habit of busting DA/DHs and MDAs.

On conventional instrument approaches, the actual statistical risk of busting descents is one of those ineffables. Where and how you do it and in what conditions drives how risky it really is. It’s one thing to burn 50 feet on a needles-centered CAT I ILS, but another to dive out the clouds from a GPS MDA with nothing showing but faith and hope. Yet pilots do it, although I suspect not routinely.

I was once at a Connecticut airport standing outside the FBO shack in dense fog with an instrument student. We were waiting to get at least a half mile to depart on a training flight. We heard one of the local jet operators check in, inbound on the ILS. I told my student he’d get to at least hear a genuine missed approach. As we stood outside on the sodden gray ramp, waiting to hear the roar of spooling engines, we heard instead the unmistakable double squeak of tires on pavement. The lineman shrugged. No way they had the required minimum vis of a mile. It was RVR weather. At least it was an empty leg.

When staring at the digital glory of synthetic vision on a PFD, the temptation to do this routinely must be overwhelming. In really low weather, the runway may not be visible out the windshield, but it’s bigger than hell right there on the PFD. What’s the risk of just going for it if you don’t see anything at DA/MDA? Probably not that much, although I doubt if there’s any meaningful data to put a number on it. Synthetic vision doesn’t depict close-in obstacles off the end of runways, but a runway with an instrument approach isn’t likely to have meaningful close-in obstacles, at least from the missed approach forward to the threshold of the runway. But this varies by approach.

I’ve flown dozens of visual approaches in airplanes with synthetic vision and never noticed one in which the runway wasn’t where it was supposed to be or was misaligned enough to detect, but I also wasn’t looking for apparent misalignment. Even the synthetic centerline matches the real world view, which is impressive when you consider that synthetic vision is doing that with WAAS GPS matched to a terrain database. We’ve come to take this level of performance for granted without thinking about how well it works. 

I haven’t tried landing under the hood with just synthetic vision because it doesn’t provide the necessary depth perception clues. But in low vis, you could certainly use it to get close enough to acquire the runway visually then land normally. I’d guess that would work in quarter-mile visibility or a little less. Are people actually doing it, despite the lack of legal framework, to bend FAR 91.175? No one I know has confessed, but human nature being human nature, I’d be surprised if someone hasn’t done it and if some aren’t doing it routinely—homegrown CAT II.

I’d like to see some system reliability and accuracy data on synthetic vision, but I think it’s time to allow its use for descent and visibility credit for not-for-hire operations. Sure, there’s risk in doing so, but let pilots assume it if they want. The fact that we’re not doing it already probably has more to do with legal inertia than technical considerations. I asked Garmin if they had synthetic vision in mind for this kind of upgrade and they declined to comment. I take that as a yes.

For years, the transport industry has used head-up displays to qualify crews for lower landing minimums or at least to smooth the transition from the gauges to the visual. I flew on in an Alaska Airlines simulator 20 years ago. I’ve poked around to find some data on how widespread HUDs are in airline use, but I can’t get an accurate sense of market penetration, but it’s probably the majority of the fleet. On a recent Southwest flight, the captain told me the entire fleet is HUD equipped and he liked it and used it on every approach, including visuals.

Periodically, we see attempts to offer HUDs for light GA aircraft but they never seem to gain traction. We reported on the latest offering in this video. MyGoFlight says it can integrate synthetic vision, but I’m not sure what that would look like.

HUDs have proven expensive and cumbersome, requiring a combiner-type display to be positioned in front of the pilot’s eyes only when needed. The military is way ahead on this technology and now projects it in a helmet display or, before that, on a mounted combiner that’s permanently in view in the pilot’s sightline out the canopy.

And for GA, how often would a HUD matter? How many pilots fly weather that requires bitter-end transition from the gauges to the visual and how much good would a HUD really do?

Not much, I ‘d guess. Or maybe not enough to justify spending $10,000 to add it. This is a rarified risk area in which there are too few accidents to make much sense of the probabilities. In the aviation press, we write evergreens with tips about how to transition from the gauges to the visual as though it’s fraught with hazards and tricks. It isn’t. Just look out the window.

The Southwest skipper told me he saw benefits in using it for visuals because it gave good windshear cures and made staying on speed easier. Fair enough. But little airplanes are still hand-eye machines and that’s why a lot of us still fly them. Do we really need or even benefit from so much technology just to land a Bonanza? Would it help?

If the FAA ever approves synthetic vision for approach mins credit, I think it would be festooned with so many system and certification requirements that it would be impossible to certify. But then at one time, I thought that about GPS approaches and non-TSO’d gyros too. Maybe the very act of my thinking this will cause it to become true. We can only wish.

Comments (21)

I have my students fly to 50 agl using the synvis during training for 2 reasons. First if they get stuck without an alternate within fuel range, then they will have gone to 50 feet and be familiar with the concept. This is only for emergency and not to be used otherwise to bust published minimums. The second reason is that human nature being what it is, some will cheat a bit to get in when wx is below minimums. I don't think this is going to embolden them, as they have already figured out that the synvis is very accurate and will "experiment" on their own in actual IMC.

I have done this many times and find the accuracy to be amazing. I also find the synvis very helpful for barely straight in approaches. Paul, you must remember the LDA 2 @ KHFD. When breaking out at minimums, the runway is somewhere on the far right side of the windshield. In summer, the trees on the dyke will help mask the REIL so it can be a bit of a challenge to find the airport. A quick look at the synvis on the G500 and there is no doubt as to where the runway is hiding. There have been a few minimums approaches where the G500 really helped knowing where to look for the runway even with being very familiar with the underlying landscape.

I have never had the cajones to go all the way to touch down, but do know that if you cut the threshold in 0-0 and hold a level pitch you will live to fly another day. Not so sure the insurance company will be thrilled with the results though.

It would be great to have the legal ability to use the synvis to continue the descent. I really don't see it happening too soon, but it may come eventually. This is another safety tool that can be had with current technology.

Posted by: Leo LeBoeuf | December 25, 2016 10:25 PM    Report this comment

If SHTF use all available resources.

If syn makes live easier, why not?

You can always disregard the syn.

If syn fails you're no worse off than if you didn't have it which isn't any worse than you're legally entitled to.

Except for the cost, why not?

Posted by: FILL CEE | December 25, 2016 11:27 PM    Report this comment

I doubt synvis will ever be allowed for approach min credit, but maybe flying another 100' lower (assuming an ILS/LPV) and halving the forward visibility. Precision approaches are easy, followed by non-precision approaches with a VDP, but plain non-precision approaches where the missed approach is right at the runway (or center of the airport, depending) are trickier.

EVS, on the other hand, might be more likely to become a legal way to count as seeing the runway environment.

Posted by: Gary Baluha | December 26, 2016 7:58 AM    Report this comment

I've always felt an obligation to teach checkride-ready instrument students zero-zero-ILS-to-a-full-stop technique - to be used only as a no-other-option lifesaving procedure. All on steam gauges. I think I'm in the minority in that regard, and I do understand why one would righteously espouse a contrary view. You need to know and trust your student's judgement. But if you don't, then why would you endorse him/her, regardless of any zero-zero training? That trust issue inevitably leads to a discussion about former-student fatalities... Never pleasant, but often useful.

Posted by: Tom Yarsley | December 26, 2016 9:06 AM    Report this comment

Paul--"We were waiting to get at least a half mile to depart on a training flight. We heard one of the local jet operators check in, inbound on the ILS. I told my student he'd get to at least hear a genuine missed approach. As we stood outside on the sodden gray ramp, waiting to hear the roar of spooling engines, we heard instead the unmistakable double squeak of tires on pavement. The lineman shrugged. No way they had the required minimum vis of a mile."

Regardless of what the tower is calling for vis, it's the FLIGHT visibility that is controlling for part 91 operators. I've landed a number of times where tower was calling it below minimums, but could easily see the approach light system--indicating I had more than the required minimums. Often, an approach like that has rollout into the fog that the tower was reporting.

I've been a long-time subscriber to sister publication IFR magazine. I can't lay my hands on the old copy, but their explanation was that if an aircraft lands safely with visibility below minimums, some towers will write it up as an "incident"--not requiring FAA notification. As I recall the article, it also had the explanation "unless some FAA puke with nothing better to do makes it a practice to go through the tower logs--in which case, tell them what you had for visibility."

Posted by: jim hanson | December 26, 2016 9:47 AM    Report this comment

"As I recall the article, it also had the explanation "unless some FAA puke with nothing better to do makes it a practice to go through the tower logs--in which case, tell them what you had for visibility."

Yeah, well...I wrote that, so I pretty much know the legalities. In the incident I cited, there was no way they had the viz. It was uniformly dense. The operator had a habit, shall I say.

Posted by: Paul Bertorelli | December 26, 2016 10:01 AM    Report this comment

With regard to a HUD for general aviation aircraft, it seems to me that the main obstacle, other than cost, is headroom space to install the projector. However, I see this as a useful application for the Google Glasses that were introduced a couple years ago. If you could connect the glasses to the SynVis through bluetooth, it would make a good HUD that could be easily put on when needed with no space limitations. Maybe a useful application for an otherwise expensive toy.

Posted by: John McNamee | December 26, 2016 11:43 AM    Report this comment

There's these guys:

Believe there was one other like it.

Posted by: Paul Bertorelli | December 26, 2016 12:41 PM    Report this comment

In the experimental world, there are guys who have put such a system together. Works well, though a bit distracting during normal flight. With some use, the distraction may go away. Just as the distractions from the PFD vanished with some use.

Posted by: Leo LeBoeuf | December 26, 2016 12:50 PM    Report this comment

I used to have students fly the ILS to the lowest possible point for two reasons. One was to show the glideslope signal is a parabola that curves upward just past the antenna and does so at a height of about 50 feet.

The second was to give the pilot a bail out in case he or she absolutely positively needed it and had no other option. The onset of wisdom has caused me to think the mental bandwidth is better devoted to encouraging judgment to avoid getting into that situation in the first place. Have a good alternate, plenty of fuel and a refined understanding that some days just aren't flyable.

And in boring down without reported minimums, the flight vis card is hardly relevant. I'm not worried about an FAA bust, but an ass bust. I worry about eroding adherence to the FARs and good operating practice to the extent that there's no restraint at all. Do that long enough, and you'll get your own personal crater. I never wanted to leave anyone with the impression they could revise the regs as they pleased.

Posted by: Paul Bertorelli | December 27, 2016 10:01 AM    Report this comment

AutoLand for light GA could be had with a snap of Garmin's corporate fingers. Cirrus should have done it with their Vision personal jet, instead of insisting on their required non-required we-didn't-test-it parachute system. No human participation; no human error. No human "cheating" or improvising; no fatigue; no rust; no get-there-itis.

Posted by: Tom Yarsley | December 27, 2016 10:44 AM    Report this comment

"No human participation; no human error. No human "cheating" or improvising; no fatigue; no rust; no get-there-otis."

Yeah, just system failure in a 10 to the minus 5 system and a giant lawsuit from a high net worth owner. Other than that, piece of cake.

Posted by: Paul Bertorelli | December 27, 2016 11:22 AM    Report this comment

Paul: Do you really believe that the 5-sigma system would fail more often than human pilots already do? Proscription of improvement because it entails change has been the FAA's paradigm since Jesus was in kindergarten. Well, at least we still have dual magnetos.

Remember your classic Boothroyd Dewhurst? Instead of killing yourself trying to improve a problematic part, can you find a way to eliminate that part? Parts not in the build never fail. Humans fail a lot. That's why eliminating them from the system holds so much opportunity for improvement.

These days, how often do you see a machinist hand-cranking a Bridgeport milling machine in a production environment? There's a reason for that.

Posted by: Tom Yarsley | December 27, 2016 12:14 PM    Report this comment

Dialogue--Jim--"As I recall the article, it also had the explanation "unless some FAA puke with nothing better to do makes it a practice to go through the tower logs--in which case, tell them what you had for visibility."

Paul--"Yeah, well...I wrote that, so I pretty much know the legalities."

Give me credit for not only being a long-time subscriber, but for recalling the text! It was a first exposure to the Bertorelli insouciancy. (laugh) Just wanted to reaffirm the fact that for Part 91 operations, the PILOT is responsible for determining controlling visibility.

Thinking about synvis credit--rather than new FAR rule-making, how about exploring adding it to the list of items that must be recognized to continue an approach (ALS, strobes, Runway end lights, terminating bar, VASI, etc)? These allow proceeding to 100 feet--where the runway is ALMOST always visible.. Perhaps "Approved and operative synthetic vision systems and training allow proceeding to a DH of 100 feet"?

That seems practical and simple, but knowing the FAA, they will require CAT II redundancy (two separate systems--a comparitor system annunciator, radar altimetry, and two pilots) making the whole process moot. I guess I like the present system better than that. As I get older, I don't want to fly much lower than 200' anyway. (sigh)

Side note to the discussion--at very low ceiling/visibility conditions, there is a REASON that touchdown and centerline lighting is required--it is difficult to judge the landing flare under those conditions.

Posted by: jim hanson | December 27, 2016 12:15 PM    Report this comment

"Do you really believe that the 5-sigma system would fail more often than human pilots already do?"

Doesn't matter what I believe. It's what the tort system can prove to a jury. And that, of course, has nothing to do with the truth. I think this alone might chill autopilot makers from pursuing it soon.

Posted by: Paul Bertorelli | December 27, 2016 12:33 PM    Report this comment

The 91.175 addition should exclude approval by the administrator. Just say "cues on a synthetic or enhanced vision that the pilot deems suitable for continuing."

Posted by: Paul Bertorelli | December 27, 2016 12:35 PM    Report this comment

I like it! By saying "that the pilot deems suitable for continuing"--it takes some burden off the manufacturer, and puts it on the pilot. Obviously, the pilot "deemed it suitable" if they elected to continue.

What a breath of fresh air--not trying to regulate for every contingency.

FAA Administrator Huerta's term is up Jan 7 2018. What would it take to pry you out of Florida and into the Washington Swamp to succeed him? (thumbs up)

Posted by: jim hanson | December 27, 2016 4:57 PM    Report this comment

Paul Bertorelli for FAA Administrator in 2018.

Posted by: Rafael Sierra | December 27, 2016 10:44 PM    Report this comment

Hey Paul,

When I was wee boy starting my aeronautical career, I worked for Boeing, where we were certifying the first HUD for Alaska Airlines. In 1980, I was a flight test engineer at Boeing field flying on board a 737 making hundreds of approaches to be able to show compliance for the FAA. Sure was weird watching the HUD from behind the pilot during the approach, and then seeing the runway show up. After 30 years, you would of thought HUDs would of made it into GA airplanes!


Posted by: Richard Mutzman | December 28, 2016 8:15 AM    Report this comment

Integrate data from short range ground sensing radar into the datastream from the synvis GPS. The radar reading would act as a check on the GPS info and add confidence that the measurement was verified by a real world reading and not just from a data file. The system would give the pilot an "abort" message if the data failed to match up within reasonable margins. If you put this on a HUD, you would have a very reliable system which would accurately "paint" the runway on the canopy. Not a bad option when the tanks are almost dry and there is no alternate in range. No reason it would not work really well, but, it would be a mite expensive for small private aircraft!

Posted by: Steven Kane | December 28, 2016 8:11 PM    Report this comment

When I got to review an early syn-vis system in 2004, I flew with it, under the hood, from takeoff to about six feet AGL on landing. Yes, I did it with an instructor along.
We agreed that it looked as though I could have done the actual landing, just by holding a steady sink rate if nothing else. (When you have a 6000 foot runway and a C-172, you don't have to smear the numbers.)
And that was pretty primitive, compared to modern equipment.
I agree with several other commenters here: if you're in trouble, do whatever it takes, whenever you must. Off-label uses may not be approved, but they may save your bacon.

Posted by: Tim Kern | January 11, 2017 9:30 AM    Report this comment

Add your comments

Log In

You must be logged in to comment

Forgot password?


Enter your information below to begin your FREE registration