Found the problem.
Does anyone know if weather balloons show up on FlightRadar24 at all? There’s something called TEST4A near Warsaw at 393ft but zero knots and aircraft type BALL. It’s symbol is a ball as well.
I’m thinking it’s purely a test, given that its ICAO code is 104A1102, which is more than 24 bits (has to be a hex code since it has an “A”, which would make it 29 bits). The fact that it’s called TEST4A and 4A appears in the fake ICAO code is also a giveaway.
I only spotted that while idly tracking RCH534 and RCH157 (both C17-A Globemaster III’s) which have been flying together North from Chania in Crete. They’ve skirted Ukraine and Belarus and are descending to land at Vilnius in Lithuania. Possibly full of equipment for Ukrainian soldiers. 
Thanks - and I’ve just spotted another weird one in the same area. 104A1100 which has plane type DRON. 349kts at 380ft. Maybe a glider?
That one has an icon that looks like a quadcopter, which would explain the DRON name. But it keeps moving in a repeating pattern, so again I think it’s just a test with fake data.
Since you folks are poking around in a war zone, I wonder if what you’re seeing is the equivalent of jamming. Pumping junk records into the public database so the adversary can’t use that as a traffic characterization or targeting tool.
I know that I don’t know. But fake targets have a long history in warfare.
Could you explain this in more detail? Where are you getting the code and how it’s decierred.
Here’s the link to the TEST4A object:
At the bottom of the sidebar is a line for “ICAO 24-BIT ADDRESS”, the code being 104A1102.
Wikipedia has an explanation of the code here:
Now, in isolation it’s hard to be sure which format the code is in: decimal, octal, binary, or hexadecimal. But having the letter A in the code means it must be hexadecimal. Plain decimal would only have the digits 0-9, and no letters. Binary is 0 or 1. Etc.
But in hexadecimal, the code must only be 6 digits long, since each digit is 4 bits and 4*6=24. It’s a 24-bit code, as per the name.
But 104A1102 is much more than 24 bits! It has 8 digits, not 6. So it must be fake somehow, and I presume from that that it’s just a test, maybe of the FlightRadar24 site, maybe something else.
:
Quoting myself from two years ago for context, and spoilered for size. Click the right-pointing triangle to see it.
As a result of the various engine failures I mentioned above which tore up the cowlings and nacelles and damaged the airplane, the NTSB a couple years ago gave a safety recommendation that the FAA research the certification standards for both engines and cowlings with an eye to updating them so these failures don’t keep happening. Plus make retrofit requirements for what are now known to be inherently inadequate designs.
EASA, the EU’s FAA-equivalent, has decided to update their regs for both engines and airframes to better address the threat of severely unbalanced non-seized engines and the damage those do to the engine surround structure.
The FAA last week announced it had decided to stand pat, saying changes to their regs were not necessary, although they intended to “pay very close attention” to this risk on new certifications. I’m reminded of a Steve Martin skit where he says somebody wouldn’t have done something bad “unless they had a darn good reason!”
NTSB took the rather unprecedented step of sending a nasty-gram to FAA HQ and to the press asking the FAA to publicly explain why it thought there wasn’t a problem whereas EASA, faced with the exact same issue, said there was a significant problem that urgently needed regulatory fixing.
Wicked burn!
We’ll see what happens.
Thanks for the detailed explanation. I’ve never used Flightradar24 much. It looks more oriented to tracking aircraft than FlightAware
Wow. That sounds like a future invitation to a Congressional hearing.
For all that we shake our heads at safety and regulatory failures in places like Ethiopia, it’s always worth remembering that to some extent, the FAA has the same problem. Albeit with a much lower recent body count on their watch.
Just for a change of pace, another accident (actually, a pair of them and very nearly a third one) that caught my attention while clicking through a bunch of aviation videos. Again, this is not new – it’s from the 90s and aviation buffs may well be familiar with them, but I wasn’t.
In March, 1991, a Boeing 737-200 operating as UAL 585 suddenly and inexplicably lost control as it was on its final approach to Colorado Springs, and dived straight into the ground. The two things that are remarkable about this event are the sudden nature of how it happened, and the fact that after 21 months of investigation, the NTSB had to close the case as “cause cannot be determined”, which is very unusual.
Then in September, 1994, a 737-300 operating as USAir 427 crashed on approach to Pittsburgh International under almost identical circumstances. In that accident, the tail section was somewhat intact and indicated full rudder deflection at the time. Also, the rudder power control unit (PCU) was intact and could be taken to the manufacturer in California for testing. It was found to be functioning perfectly, passing all tests. There was increasing suspicion that the PCU was somehow the problem, but nothing conclusive could be established.
In June, 1996, while the USAir investigation was still ongoing, a 737-200 operating as Eastwing 517 again experienced the same loss of control on approach to Richmond, Virginia. This time, however, the problem corrected itself, happened a second time, and corrected itself again. The pilots were well aware that if the problem recurred at low altitude no recovery would be possible. Fortunately, it did not, and they landed safely.
The NTSB now had an intact airplane to examine and surviving pilots to interview. Once again, the rudder PCU was the suspect, and once again it was taken to the manufacturer, Parker Hannfin in California, for testing. And once again, it passed all tests. It was working perfectly.
Then someone had a brilliant idea. What if the PCU had been subject to thermal shock? They cooled the test PCU down to low temperatures with dry ice, then operated it with hot hydraulic fluid. Before long, one of the actuators jammed.
But it got more interesting. A Boeing engineer informed the NTSB that the design of the dual servo valves was such that if one of the actuator slides jammed, the PCU would reverse its function and move the rudder in the opposite direction to that commanded by the pilot.
The end result of all this is that the NTSB was able to go back and ascribe conclusive causes to all three events. It was fortunate that the Eastwing flight landed safely and a cause could be determined as another crash would likely have grounded the entire 737 fleet worldwide. In the aftermath, Boeing redesigned the PCU and retrofitted it to all 737s in service, and pilots were given additional training to deal with similar anomalies.
This Wiki entry gives some of the facts but lacks many of the details.
Just to add, although the 90s may seem like ancient history to many of you young’uns, it happened to be a period in my career when I was doing a great deal of flying on business – typically several flights a week, all over the continent. And I’m sure much of it on the then-defective Boeing 737s. 
ETA: I know that the 737 has a well-deserved reputation as one of the safest and most popular airliners ever produced. But still, there have been these specific problems …
Excellent post, thank you.
The perfect aeroplane has never been (and will never be) built - there are just too many confounding factors. The best we can hope for is not to be one of the ‘test passengers’ when the design flaw manifests itself. Thanks to decades of expertise and hard learning experiences, the odds are well in our favour - and in fact have been since the start of air passenger travel, but overwhelmingly so now, especially when it comes to North American, European, Australian, and New Zealand airlines.
One of the things that came out of those 737 mishaps (and other contemporary mishaps) was a call for much more capable flight data recorders. As airplanes were becoming increasingly digital anyhow, this wasn’t too difficult. But back in those early days, attitude, altitude, airspeed, heading, and engine power were about all that was recorded. That plus scratchy crappy audio and whatever hunks of wreckage they could retrieve and analyse. Accident investigation had a lot of guesswork about the “what”, much less the harder “why” and “how” which are needed to prevent recurrences.
We’ve come a long way, baby.
The aftermath is perhaps interesting too.
After they discovered the central problem with the PCU, there was an unusually long process of sorting out how to fix it. Despite the crashes, that particular failure rate was so infinitesimal that a complete redesign had a higher risk of introducing a different and more frequent failure mode. It was broke, but the fix might easily have made things worse; differently worse, but worse nonetheless. So they had to be very, very sure their improvement was really an improvement.
Eventually they came up with a minimally intrusive way to tweak the problem away, and also to include a system to detect and override the result of any reversal no matter what malfunction may have caused it.
AFAIK there have been no rudder issues w the 737 since.
Different topic:
I happened to be looking for an old thread about something unrelated and came upon the one about the Air Canada A320 that tried to land on the taxiway at SFO a few years ago. A disaster was narrowly averted when they went around instead. NTSB Investigation Docket in case anyone is interested in the details.
The point here today is the NTSB’s findings said there were significant problems with how both Air Canada’s systems and their pilots processed NOTAM information. As a result the pilots were unaware of the NOTAMed runway closure which led them to misidentify which row of which lights was connected to which strip of concrete on the ground. Oops.
This points out what’s at stake when pilots don’t have, or don’t process and understand, all the relevant unpleasant surprises that NOTAMs represent.
The recent FAA NOTAM system failure and resulting flight cancellation exercise makes this topic suddenly timely again.
It surprises me to hear that, as everything I’ve read over the years has suggested that Air Canada’s procedures were exceptionally rigorous, and certainly their safety record is very good. I remember years ago Lockheed praising Air Canada’s rigorous evaluation methodology when they selected the L-1011 out of competing bids – obviously praise motivated by self-interest, but likely reflecting an actual reality. And of course that other jumbo tri-jet, the DC-10, went on to become rather infamous for all its problems.
But then again, it’s equally surprising that this near-catastrophe could possibly have happened, and yet it did. And catastrophe it would have been – there were four aircraft on that taxiway, three of them jumbos.
As an aside, I’m endlessly impressed by the coolness of the exchange between the tower and ACA 759. This was, as near as I can recall, the exchange that occurred seconds before a potential catastrophic disaster:
Tower: Air Canada 759, go around.
ACA759: Air Canada 759, going around.
Cool as a cucumber! All the drama came later.
If that can’t tell the difference between a runway and a taxiway then the NOTAM isn’t going to help.
But your general point regarding NOTAM’s is true.