Hi, I’m Bruce Landsberg and welcome to the Leading Edge. We’ll discuss safety-of-flight issues, procedures, techniques, and judgment. With the convective nature of Internet misinformation, and so much content that is over weight and out of balance, you need an experienced and trusted source. So, strap in and let’s go fly!

The NOTAM System made me do it!

August 10, 2015 by Bruce Landsberg

stack of papers  b&wAn amazing bit of radio chatter played out on national news recently when an airline pilot headed to Fargo, North Dakota, implied he had only five minutes of fuel left. As bad luck would have it, the Blue Angels were practicing at the airport at that very moment.

The controller advised that there would be a 20 minute delay! Huh? Doing the math, the 15 minute deficit implied the MD80 would plunk down in the prairie but with no fire and the Angels would complete their practice on schedule. The pilot finally made it clear that he was coming in to land and they’d sort it out on the ground. Good call. The captain properly exercised PIC authority to get the aircraft on the ground safely. In too many cases GA pilots abdicate responsibility when they get into difficulty, being more afraid of legalities than gravity. As shown in comments, what the Captain meant to say,  was that in five minutes he would  start burning into his reserve fuel – The proper terminology would have been  a “minimum fuel advisory” as opposed to the military term of “Bingo” fuel which is not in the lexicon of civilian aviation.

A pertinent regulation, if I have it correctly, is FAR 121.639—Fuel supply: All domestic operations. No person may dispatch or take off an airplane unless it has enough fuel—(a) To fly to the airport to which it is dispatched; (b) Thereafter, to fly to and land at the most distant alternate airport (where required) for the airport to which dispatched; and (c) Thereafter, to fly for 45 minutes at normal cruising fuel consumption.

FAR 121.647 Factors for computing fuel required. Each person computing fuel required for the purposes of this subpart shall consider the following: (a) Wind and other weather conditions forecast. (b) Anticipated traffic delays. (c) One instrument approach and possible missed approach at destination. (d) Any other conditions that may delay landing of the aircraft.

So what did the crew and the dispatcher know, and when did they know it? Let’s acknowledge that mistakes were made in checking notams—which would have noted the closure times—and that perhaps just enough fuel was on board to be legal, or not. And let’s also acknowledge that the current notam system is a mess, and has been for years. It is a convenient place to dump everything—much of it not operationally pertinent…and some of it critical! Apparently, a lot of pilots and even a few controllers are getting caught in the notam swamp, as the NASA’s Aviation Safety Reporting System just noted.

Over a decade ago, the FAA promised that the system would be made more relevant and one could filter the data to what was important to your flight operation. No results! After passage of the Pilot’s Bill of Rights (PBOR) in 2012 in Congress, another committee was formed to address the unworkable. From the bill, “...The goals of the NOTAM Improvement Program are—

(1) to decrease the overwhelming volume of NOTAMs an airman receives when retrieving airman information prior to a flight in the national airspace system;

(2) make the NOTAMs more specific and relevant to the airman’s route and in a format that is more useable to the airman;

(3) to provide a full set of NOTAM results in addition to specific information requested by airmen;

(4) to provide a document that is easily searchable; and (

5) to provide a filtering mechanism similar to that provided by the Department of Defense Notices to Airmen.”

This was supposed to be completed in a year but things slipped a bit. If Google can manage an infinite number of possibilities, it would seem the FAA could at least do better than the current system with a much more limited data set. Some of the aftermarket providers do a better job of filtering but why not the official source?

Flying a fixed wing aircraft in daylight hours, why would I care that a tower light was out on an obstacle five miles away from the airport and only 198′ agl? I wouldn’t, but an EMS helicopter crew on a night mission might. The system is also often geographically confused. A flight from Maryland to South Carolina will list items like turbine wind farms in New York, runway restrictions in Georgia, obstacle unlightings more than 75 miles off to one side, or laser light shows in faraway places. On a recent DUAT Briefing, I lost count with more than 400 notams listed for a 500 mile trip: Must be a rule of one notam per mile flown.

Airport/runway closures and TFRs along with non-availability of certain IFR approaches ought to be bold printed in red. And is it really necessary to use arcane abbreviations and acronyms in today’s deep bandwidth environment? Fargo’s closure somehow slipped through the cracks.

The FAA could use this as an opportunity to 1) rise above the punitive mentality that periodically pervades the agency, and 2) look at the root cause of this incident. My bet is that this crew and dispatcher will be very careful to check airport status going forward. Since the FAA is more than two years behind addressing the notam system as specified in the PBOR, how about suspending penalties—except in cases of willful disobedience—until the new system is operational? It should also recognize that a poorly executed notam system is more than an embarrassment—it is a critical safety issue.

The FAA’s new system is planned for roll out at year’s end—that should make things much easier and safer, we’ll see.

Has anyone else missed something in the notam system and how did it affect your flight?

Bruce Landsberg,
Senior Safety Advisor, Air Safety Institute

ASI Online Safety Courses  |  ASI Safety Quiz

Velociraptors in the air

July 22, 2015 by Bruce Landsberg

344523_AOPAThey’re out there waiting for us. Lurking, often invisible, and then pouncing with devastating swiftness. While V-Raptors haven’t been seen on the planet for a few million years, the collision possibility with another aircraft will have much the same effect (although one usually isn’t eaten after the encounter).

The Wright brothers had no worries about midair collisions until the second aircraft was introduced and we’ve had to be on guard ever since. Last week, an F-16 and a Cessna 150 collided between 1,500′ and 2,000′ agl near a small nontowered airport in South Carolina, with tragic consequences.

The jet, based at Shaw AFB just up the road to the north, was practicing instrument approaches at nearby joint-use Charleston AFB/International Airport (CHS). It’s home to a USAF C-17 wing, a moderately busy airline terminal, and the usual mix of light and heavy GA aircraft. The weather was VFR so primary separation was based on the principle of see and avoid.

From the NTSB’s preliminary report: “At 1057:41, a radar target displaying a visual flight rules transponder code of 1200, and later correlated to be the accident Cessna, appeared in the vicinity of the departure end of runway 23 at MKS [Berkeley County Airport] , at an indicated altitude of 200 feet. The Cessna continued its climb, and began tracking generally southeast over the next 3 minutes. For the duration of its flight, the pilot of the Cessna did not contact CHS approach control, nor was he required to do so. At 1100:18, the controller advised the pilot of the F-16, ‘traffic 12 o’clock, 2 miles, opposite direction, 1,200 [feet altitude] indicated, type unknown.’ The F-16 pilot responded and advised the controller that he was ‘looking’ for the traffic. At 1100:26, the controller advised the F-16 pilot, ‘turn left heading 180 if you don’t have that traffic in sight.’ The pilot responded by asking, ‘confirm 2 miles?’ Eight seconds later, the controller stated, ‘if you don’t have that traffic in sight turn left heading 180 immediately.’ Over the next 18 seconds, the track of the F-16 began turning southerly.” The collision occurred shortly after that—there will be more precise information in the final report.

What about see and avoid? To avoid, one has to see first. In some circumstances it works exceptionally well but in others, not so much. As always, there are multiple factors that become ever so clear in hindsight. Cessna 150s are not known for stellar cockpit visibility despite many other excellent attributes. However, F-16s were designed with the fighter pilot’s adage in mind—lose sight and lose the fight. But no matter what we’re flying, if we’re not looking at the right place at the right time, something could bite us faster than the feared velociraptor in Jurassic Park—especially with high speeds involved.

344523_AOPAA nasty reality of physics is that objects on a collision course will not appear to move. They stay small and unobtrusive, sometimes visible to the pilot and sometimes hidden. Then, suddenly, the intruder aircraft explodes into full view growing very quickly, but by then it’s often too late to avoid.

With a big sky and not many aircraft, the odds of a collision on any day in any location are small—but obviously, where traffic congregates the potential goes up. GA-on-GA collisions are largely within several miles of an airport in good weather. Typically, there are between five and 10 collisions annually. The numbers have been gradually declining which could be due to reduced flight activity, more aircraft being equipped with traffic advisory equipment, or more likely a combination of both. It’s not because pilots have become better traffic spotters.

Military-on-GA collisions are rare, thankfully, due to the efforts to keep the fast movers in mostly sanitized airspace such as restricted areas or Military Operations Areas (MOAs). However, in this case, the jet was down low and the pilot was likely very busy getting set up for the approach—we’ll know more when the investigation is complete. Most military fighters have to move fast in the pattern because they may get a bit wobbly below 200 knots.

There hasn’t been a midair collision between GA and an air carrier since a Piper Cherokee smacked a DC-9 over Cerritos, California, in 1986. That led to the requirement for Mode C (altitude reporting) transponders for anyone flying in the airspace listed below (with a few exceptions) and the development of Traffic Collision Avoidance Systems (TCAS) for all airliners:

  • Operations within Class A, Class B, and Class C airspace.
  • Operations within 30 nautical miles of the primary airport within Class B airspace from the surface to 10,000 feet msl.
  • Operations above the ceiling and within the lateral boundaries of Class B and C airspace.
  • Operations above 10,000 feet msl in the contiguous 48 states, excluding the airspace at and below 2,500 feet agl.

TCAS and Mode C essentially resolved the air carrier collision problem but it cost everyone a lot of money.

The FAA’s new mandate requiring most aircraft to be equipped with Automatic Dependent Surveillance-Broadcast (ADS-B) “out” equipment will take effect in about five years. (Learn about AOPA’s position on ADS-B in this Air Traffic Services Brief on ADS-B, and learn more about ADS-B technology in AOPA’s online resources page.) It’s not cheap, but then neither were Mode C transponders, which still must be carried because of the TCAS gear on airliners.

I’m conflicted on several things here: The FAA will save hundreds of millions of dollars annually by decommissioning a lot of radars, but it will provide no help on equipage for the airlines, GA, or the military. There is no relief from having to carry Mode C, and at this writing the FAA requires a certificated “out” unit to get the benefits of seeing traffic and weather uplinked to any aircraft if equipped with ADS-B “in” equipment. Portable and non-certificated gear (much less expensive) need not apply. Maybe it’s time to rethink that.

Having the good fortune to fly with a certificated unit (bit the bullet this spring) I see all the traffic that ATC sees—essentially “built-in flight following” that is not controller-workload dependent. It allows to me to visually acquire MUCH more traffic than I spot using the famed Mark VIII eyeball. It’s not perfect but it’s so much better than any pilot’s scan pattern—especially mine. In this case, I’ll place my bet on the technology every time.

Mind the V-Raptors—they’re out there…

Additional Air Safety Institute resources for collision avoidance:

Collision Avoidance Strategies and Tactics Safety Advisor

Operations at Nontowered Airports Safety Advisor

Collision Avoidance Safety Spotlight

Bruce Landsberg,
Senior Safety Advisor, Air Safety Institute

ASI Online Safety Courses  |  ASI Safety Quiz

Magic and messes

July 8, 2015 by Bruce Landsberg

Perhaps you have a friend like this, a certifiable professional geek who helps to manage the technical (read: computer) complexity that’s become a way of life. My “smart” phone reverted back a few generations and wasn’t behaving as intelligently as usual. Google wisdom offered, “It does that and if you reboot, the problem usually goes away.”

My friend had a simpler strategic explanation:

“Welcome to my world of buggy software. This stuff is so complex any more it is almost beyond human comprehension. They modularize the system as much as possible with the published system calls  ‘supported between modules’ and hope it all works together correctly. And testing only confirms the presence of defects, never the absence of defects…

You can always find people who understand how a particular system module works, or even a sub-system (a collection of a few modules, e.g., memory management) but the WHOLE system…not really.”

We are seeing increasing levels of software in our aviation hardware, and most of the time it works beautifully, but sometimes it’s a bridge too far in my opinion. A few points for your consideration as we look to the future:

In May, an Airbus A400M 4-engine turboprop (similar to a C-130) crashed on takeoff when 3 of the 4 engines suffered what Airbus called “power frozen” after lift-off, and failed to respond to the crew’s efforts to regain control.

A British newsletter, The Register, noted:

People familiar with the investigation said the torque calibration parameters for the engines were wiped during the installation. This data is needed to measure and interpret information coming back from the A400M’s engines, and is crucial for the Electronic Control Units (ECU) that control the aircraft’s power systems.

Without that sensor data, the ECU automatically shut down the engines, or at least put them into the lowest power settings. According to safety documentation, the pilots would only get a warning from the ECUs when the aircraft is 400 feet (120 meters) off the ground.”  

Not exactly inspiring confidence, and they don’t fly well on one engine apparently.

From Fortune magazine:

United Airlines unveiled a program designed to award free frequent flier miles to potential hackers who could break into the company’s mobile applications and websites. The company said that the program would not be open to hackers wanting to crack into an airplane’s Wi-Fi or on-board entertainment, or control systems…

According to the Wall Street Journal report, Boeing has also turned to outside security experts to uncover security bugs. As part of a security program, the airline maker is paying friendly hackers to break into the onboard software of its 787 Dreamliner.” 

So the theory is that if a hacker has enough confidence in the onboard aircraft systems to cash in his frequent flyer miles, it’s good enough for the friendly sky folks. But don’t mess with the entertainment stuff—that’s where the real money is!

From the NY Times a few months back:

“Federal regulators will order operators of Boeing 787 Dreamliners to shut down the plane’s electrical power periodically after Boeing discovered a software error that could result in a total loss of power.” 

Sort of similar to my cell phone – reboot and your problems will go away.

Just this morning, CNN reported that United Airlines was handwriting tickets as they grounded all flights. CNN said, “United spokesmen were not available for comment, but the airline’s Twitter account responded to customer complaints, saying, ‘We are working on getting you to your destination as quickly as possible.’”

Hmmm…this happened over the winter as well. Must be a bad module—again.

I don’t mind an occasional blank avionics screen, especially if there is redundancy, but the engine(s) and flight control systems need to be 100%, or at least 99.9999%. For GA aircraft there are unconfirmed rumors that the brightest minds are thinking about fly-by-wire. If you’ll pardon the pun, we’ve had it for a century—good old stainless steel wires (or occasionally pushrods) that work beautifully well, is stone simple, and relatively cheap.  Don’t need to hire hackers to try to break the system. Don’t need to reboot it or wonder what it’s going to do next. It just works. It does need some adjustment occasionally but if reasonably maintained, will not fail catastrophically due to a module mix-up or a packet going POOF. Any A & P can fix/maintain it.

As noted by United’s latest SNAFU, the problem is becoming more pervasive—can you say, U.S. Government hacks?

But, back to our corner of the aviation world, I like simplicity—especially when it works. More isn’t better! I’d rather the best government, academic, and industry minds come up with an economic/engineering model to build quality light aircraft in affordable quantities so the companies can make some money and pilots can afford to buy and operate new ones. Maybe that makes software complexity look like child’s play, but it’s something the industry could certainly use.

What do you think?

Bruce Landsberg,
Senior Safety Advisor, Air Safety Institute

ASI Online Safety Courses  |  ASI Safety Quiz