How can certified avionics be so crappy?

[MountainDude]

My understanding was that certified equipment is so expensive because it has to be super robust.
Then you read about G5s and GPS175s not working and Garmin not being helpful, chronic issues with current BK devices they can't figure out how to fix, and AV-30 from uAvionix having lots of issues. Macs and iPhones are more complicated, yet they simply work, even though they are not certified. Why cant avionics work as robustly as iDevices, or Dell desktops? Very frustrating.

 

[eman1200]

 

[TCABM]

Ever try to hook a fax machine to your cell phone with an RS-232 cable?

 

[Clip4]

My understanding was that certified equipment is so expensive because it has to be super robust.
Then you read about G5s and GPS175s not working and Garmin not being helpful, chronic issues with current BK devices they can't figure out how to fix, and AV-30 from uAvionix having lots of issues. Macs and iPhones are more complicated, yet they simply work, even though they are not certified. Why cant avionics work as robustly as iDevices, or Dell desktops? Very frustrating.

I work with a fleet of 12 G5 equipped trainers flying an average of 1000 hours a year each and 3 have GPS 175s. Problems are rare.

 

[machkhatib]

Problems are rare.

As the guy who has a G5 problem, I actually agree with this statement. Garmin letting a lemon slip through to the consumer isn’t the reason for my anger - it happens. Rather, I’m angry at Garmin for resisting for months to replace a G5 under warranty.

https://pilotsofamerica.com/community/threads/garmin-warranty.146190/

 

[wanttaja]

I was once shopping for a new radio for my car. The salesman told me that one particular model was "aircraft quality." He didn't understand why I burst out laughing.....

Ron Wanttaja

 

[MooneyDriver78]

It’s not certified to be robust, but it follows standards, functionality and no emissions that could interfere with radios (That’s why you need to turn off electronics when landing in commercial aircraft).

Also don’t forget that there are more iPhones sold in the 1st week than Garmins G5s in its entirety. So prices of our avionics are going to always be high for what you get. And the manufacturing process probably has more humans and human errors.

 

[FORANE]

I was once shopping for a new radio for my car. The salesman told me that one particular model was "aircraft quality." He didn't understand why I burst out laughing.....

Ron Wanttaja

When I was looking at houses to purchase a listing agent stated the pool cover was strong enough to walk on. He proceeded to walk on it and fall in the pool, in the winter...

 

[bnt83]

My understanding was that certified equipment is so expensive because it has to be super robust.
Then you read about G5s and GPS175s not working and Garmin not being helpful, chronic issues with current BK devices they can't figure out how to fix, and AV-30 from uAvionix having lots of issues. Macs and iPhones are more complicated, yet they simply work, even though they are not certified. Why cant avionics work as robustly as iDevices, or Dell desktops? Very frustrating.

Automotive world is 10x worse. Home appliances are 10x worse.

Dell? I've never had multiple power supplies fail in my avionics.

 

[MountainDude]

Automotive world is 10x worse. Home appliances are 10x worse.

Dell? I've never had multiple power supplies fail in my avionics.

None of those things require certification that is "super expensive and time-consuming".

 

[TCABM]

My understanding was that certified equipment is so expensive because it has to be super robust.

That’s not a sound understanding.

Certification is expensive because it requires documented compliance with a particular technical standard. In some instances this may be down to an individual model of aircraft, by year group and/or serial number. It costs real dollars to document compliance and those costs are baked into a product for a relatively small market.

Attaching certification to any particular aviation item also attaches a degree of liability to the item owner. This degree of liability is measured in dollars and, again, that cost is also baked into product pricing for a relatively small market.

ETA: as to the automotive market, requirements and standards exist there, too. Much of the cost component is recovered thru volume sales. FrEx, there were 15.5M new cars sold in the US last year. There were 1,135 new piston aircraft sold.

GAMA Releases Third Quarter 2023 Shipment and Billing Report - GAMA

FOR IMMEDIATE RELEASE {gama.date_MDY_formated} Contact: Andre Castro GAMA Releases Third Quarter 2023 Shipment and Billing Report WASHINGTON, D.C....

gama.aero

 

[robin ardoin]

Bad news travels twice as fast and ten times further than good news. It is much more common for a dissatisfied consumer to take the time to post online about a problem they are encountering with a product versus a completely satisfied consumer. Understandably so, as the person with a problem will be more likely to post to seek advice/air complaints in hope of finding a solution or just venting frustration. I know of no users of G5's in my circle that have had issues with the installation or long term function. However, if I suffered the problems of the OP I would seek advice from online sources if Garmin was not responding to my/the shop's satisfaction. I think overall Garmin certified products are VERY robust but installation challenges in retrofits and the possible lemon are challenges any brand will encounter.

 

[Gary Ward]

IMO the Garmin hardware is robust.
But many times the installation is not and Garmin get's blamed.

 

[mondtster]

None of those things require certification that is "super expensive and time-consuming".

That is untrue in the automotive world. People just don’t understand what is done and how it is done. I work for an OEM and I’d bet our expenditures are much higher and more time consuming.

Ever had an emissions check done on a car?

 

[Bell206]

My understanding was that certified equipment is so expensive because it has to be super robust.

FYI: Certified aircraft equipment simply means the item was built and tested to a known standard: TSO, RCTA, SAE, etc. Those standards include various environmental requirements for EMI, moisture, and so on depending on the equipment. No where does it mean it equal "super robust" at least in my experience. For example, when Garmin makes a certified and non-certified version of the same equipment, they both are built on the same production line except at the end of the line one gets it approval mark and the other doesn't.

 

[morleyz]

My understanding was that certified equipment is so expensive because it has to be super robust.
Then you read about G5s and GPS175s not working and Garmin not being helpful, chronic issues with current BK devices they can't figure out how to fix, and AV-30 from uAvionix having lots of issues. Macs and iPhones are more complicated, yet they simply work, even though they are not certified. Why cant avionics work as robustly as iDevices, or Dell desktops? Very frustrating.

I've certainly had my iPad shut off due to heat and ForeFlight crash in flight, so I wouldn't exactly say they're without issues. As someone who has been in IT for 30+ years, I wouldn't say typical computer products are particularly robust. Now the interfaces on avionics, automotive, smart devices, etc...that is something that could use addressing. Clearing the developers and designers are not real users.

 

[asicer]

Ever had an emissions check done on a car?

Plug in an OBD2 scanner, click "upload" and check that the catalytic converter hasn't been stolen?

 

[mondtster]

Plug in an OBD2 scanner, click "upload" and check that the catalytic converter hasn't been stolen?

My point was that the OP suggested that there wasn’t as extensive certification done on automobiles. There most certainly is, and it costs the OEMs millions of dollars. Some areas of the US also expect continued compliance with the standards that the OEMs are held to via emissions checks and safety inspections.

At least there isn’t government oversight of the aftermarket replacement parts industry like there is in aviation. Parts would be even more expensive when your converter is stolen.

 

[MountainDude]

My point was that the OP suggested that there wasn’t as extensive certification done on automobiles. There most certainly is, and it costs the OEMs millions of dollars. Some areas of the US also expect continued compliance with the standards that the OEMs are held to via emissions checks and safety inspections.

At least there isn’t government oversight of the aftermarket replacement parts industry like there is in aviation. Parts would be even more expensive when your converter is stolen.

Sorry if I was not clear. I do not know anything about auto certifications. However, that is not relevant. The part that is frustrating is that the avionics are held to a very high standard, yet break down often or have an inherent flaw in their design or manufacturing.

 

[MountainDude]

That’s not a sound understanding.

Certification is expensive because it requires documented compliance with a particular technical standard. In some instances this may be down to an individual model of aircraft, by year group and/or serial number. It costs real dollars to document compliance and those costs are baked into a product for a relatively small market.

Attaching certification to any particular aviation item also attaches a degree of liability to the item owner. This degree of liability is measured in dollars and, again, that cost is also baked into product pricing for a relatively small market.

ETA: as to the automotive market, requirements and standards exist there, too. Much of the cost component is recovered thru volume sales. FrEx, there were 15.5M new cars sold in the US last year. There were 1,135 new piston aircraft sold.

GAMA Releases Third Quarter 2023 Shipment and Billing Report - GAMA

FOR IMMEDIATE RELEASE {gama.date_MDY_formated} Contact: Andre Castro GAMA Releases Third Quarter 2023 Shipment and Billing Report WASHINGTON, D.C....

gama.aero

Thank you for the explanation. I understand the cost-per-unit advantage of the auto parts and consumer electronics. Let's not focus on the costs. My understanding had to do with the avionics that is certified to not fail and endanger flights, especially in IMC, yet there are so many reports of failures, some being pretty systematic (like AV-30).

 

[mondtster]

Sorry if I was not clear. I do not know anything about auto certifications. However, that is not relevant. The part that is frustrating is that the avionics are held to a very high standard, yet break down often or have an inherent flaw in their design or manufacturing.

In my opinion, it is still relevant. Testing and certification costs in the automotive industry are high and thanks to trying to meet the increasingly difficult regulations while building to an affordable price point the durability/reliability of new vehicles and their components are slipping. Outside of installer induced problems with your avionics, I'd be expecting to visit a car dealership for various problems with a higher frequency than I would a radio shop.

But it is also important to have realistic expectations. Aircraft owners seem to have the unrealistic idea that Garmin should continue to support 20 year old avionics, which likely contain parts that are no longer available thanks to the ever evolving consumer electronics industry. Car owners don't seem to place the same expectations on the OEMs, nor is the support really provided.

 

[Daleandee]

Aircraft owners seem to have the unrealistic idea that Garmin should continue to support 20 year old avionics, which likely contain parts that are no longer available thanks to the ever evolving consumer electronics industry. Car owners don't seem to place the same expectations on the OEMs, nor is the support really provided.

Easy to find aftermarket replacements parts for my 50 year old car. With 20 year old certified stuff it's quite a different story ...

 

[Gary Ward]

That is untrue in the automotive world. People just don’t understand what is done and how it is done. I work for an OEM and I’d bet our expenditures are much higher and more time consuming.

Ever had an emissions check done on a car?

Yea years ago we had a smog check here in town. They built numerous nice 4-5 bay brick buildings all around town to check our tailpipes.
I went to one with my 1971 dodge challenger running a big block with a radical racing cam, hi compression pistons, headers, running on straight 110 octane heavily leaded racing gas stinking like crazy because of the overlap in the cam. It passed easily! I didn't think much of it after that.
Not too long after that they did away with smog check in our area and those nice brick buildings sat vacate for years. What a waste of money and everyone's time.

 

[TCABM]

… My understanding had to do with the avionics that is certified to not fail and endanger flights, especially in IMC…

Again, your understanding is incorrect. Talking about the AV-30, it’s an EFIS, and it’s governed under TSO-C209. You can read it at the link below, it’s only 6 pages.

Dynamic Regulatory System

drs.faa.gov

The relevant paragraph is

Failure Condition Classifications.
There is no standard minimum failure condition classification for this TSO. The failure condition classification appropriate for the equipment will depend on the intended use of the equipment in a specific aircraft. Document the loss of function and malfunction failure condition classification for which the equipment is designed.

Further, the doc says refer to an RTCA document (RTCA/DO-178C) for software qualification. That doc states failures are classified in 5 categories, from no safety effect to catastrophic.

I’m not familiar with the particulars of AV-30 problems, but I’d hazard to guess the manufacturer is, has documented them, has classified them, and has provided guidance to users on limitations the user has to implement on some of the more hazardous failure classifications.

But that’s just the instrument. The installer is making his own wiring harness, and there’s no real controls in place if he doesn’t pin out something correctly, doesn’t use the specified equipment and supplies when making this harness, or doesn’t install it correctly.

Also, the EFIS may not be the actual source of the problem; it may be in a separate component that may be failing.

 

[Bell206]

The part that is frustrating is that the avionics are held to a very high standard, yet break down often or have an inherent flaw in their design or manufacturing.

The standards under which this equipment is held to and certified are basically a minimum standard vs a "higher" standard. Some vendors go beyond and some don't for various reasons. However, if you want a certified lower failure rate then you need to buy at the higher levels where failure rates are part of the process.

My understanding had to do with the avionics that is certified to not fail and endanger flights, especially in IMC, yet there are so many reports of failures, some being pretty systematic (like AV-30).

FYI: if you dig deeper you'll find a majority of avionics "failures" are more installation related than the specific equipment related. However, if any of these type failures did in fact "endanger" flights due to an "inherent flaw", rest assured there would be an AD in the making or issued to correct the problem.

Easy to find aftermarket replacements parts for my 50 year old car. With 20 year old certified stuff it's quite a different story ...

Not really. You just need to follow a different process and find someone to sign off on the replacement part. In general, its somewhat similar to a vehicle factory dealer won't install (sign-off) aftermarket parts at the dealership as they would have to provide a warranty in most cases. But depending on what you wanted to replace on your aircraft you might be surprised at your options.

 

[Pinecone]

Avionics are also expensive due to the low volume

In 2022 about 2.86 MILLION cars were sold. In ONE year.

There are about 204,000 GA aircraft in the US. Total.

 

[Deelee]

View attachment 125631

This gives me severe ptsd. We dealt with the dreaded double red Xs a few years ago with our Aspen. All fixed now, but whenever I see a big red X, all I see is $$$$$$$$$$$$$$$$$$$

 

[AV8R_87]

TSO-C209, DO-160, DO-178, DO-254. Each with their own reference documents, processes and procedures to follow. Functional Hazard Assesments, System Safety Assessments, Failure Modes And Effects Analysis. And there's a lot more involved in the process of certification. Then add some money for liability.

Easy to see the difference between the certified AV-30 and the experimental one.

 

[hindsight2020]

Easy to see the difference between the certified AV-30 and the experimental one.

They're the same picture...

 

[AV8R_87]

Not the same price, for the reasons quoted above.
Also, that's why the E version software takes six months to make it to the C version. Takes that long to get through all the certification requirements.

 

[hindsight2020]

Not the same price, for the reasons quoted above.
Also, that's why the E version software takes six months to make it to the C version. Takes that long to get through all the certification requirements.

You missed my point. The MTBF between the two boxes is identical, which is what people in the end care about. If they weren't (substantively), experimental owners would be slapping C versions all over their E rigs. Or put conversely, the FAA would ban the use of the E version in class ABCD...(maybe even E). They don't, they let these yahoos fly behind an ipad AHRS-emulator over the same ground track a 777 with 300 souls on board is bound to pass 3 minutes later, to say nothing of the non-participating bystanders on the ground. Single standard of safety alright....

BL, it's all consumer grade junk in the class-I space from where I sit as someone who (stupidly, username checks) flies for money. I don't mind the fact it's [comparative] junk... I mind the premium.

In fairness, the last thing I'd want down here in small fry land is the cost impositions of part 25 aircraft avionics. I don't have a boat so I can't speak how robust those doo dads are by comparison. I know I'm pretty happy with the electronics in my lower-middle class Hyundai, so there's that. People swear all the connectors and display would instantly melt the second I would put wings on it, but that's as always, being asked to prove a negative.

What I also know is that the analog avionics in my former plane were pretty craptastic. The 430 otoh, buttons were starting to take some effort to work, but the screen or power supply never died... before the engine did. Granted, I don't think it was worth 8 grand before installation either, but I'm not fundamentally opposed to electronics in the cockpit (an oft-mischaracterization of my position on this topic).

 

[Bob Noel]

I don't recall seeing any MTBF specifications in any TSO or RTCA or the like. People seem to think that certified means it never fails. From one aspect, It just can't fail in a hazardous manner.

wrt certified costs vs uncertified - there are companies that know how to develop software to comply with certification requirements without adding substantial costs. Other companies apply certification tasks as adjunct work, rather than baking it into the overall process. For those companies, of course certified means higher costs.

Years ago NASA and the FAA did a workshop series (Streamlining Software Aspects of Certification). They were looking at cost drivers because of certification requirements. I don't know if it made it into the final report, but during side discussions with (I'm not going to name-drop) it was observed that the companies that knew how to develop software efficiently weren't particularly interested in sharing their know-how with competitors... and the FAA personnel can't divulge that information either.

 

[AV8R_87]

The MTBF between the two boxes is identical

Chicken and egg argument. The experimental box is so reliable because the certification standards and requirements are recognized to be a good thing and those design methods are implemented for the experimental stuff. AV-30 example aside, where the hardware is identical, with qualification requirements driven by the certified version. Makes no economic sense to develop different hardware for the E version, but the certification costs won't be passed to the E customers, but only to the C customers. And trust me, DO-160 lab testing isn't cheap. Especially when you fail and have to redesign the hardware.
Same with the software. It's one thing to design and write software using the same practices required for certified equipment, but validation and documentation (plus certification) adds another level of cost that won't be passed to the E customers.
Good software validation tools cost a lot. If I don't have to show compliance to DO-178 I won't buy them, I will use a much cheaper internal process that works (probably) just as good. But then I don't get to put a "DO-178C Level B" sticker on that box, which means I can't use it in "real" airplanes.

I don't recall seeing any MTBF specifications in any TSO or RTCA or the like.

SAE ARP4761 defines failure categories and their associated MTBF requirements.

Hazard Classification	Development Assurance Level	Maximum Probability per Flight Hour
Catastrophic	A	10−9
Hazardous	B	10−7
Major	C	10−5
Minor	D	10−3
No Effect	E	--

Failure Condition Category	Description	Software Development Assurance Level
Catastrophic	Failure conditions that would prevent continued safe flight and landing	A
Hazardous/ Severe Major	Failure Conditions that would reduce the capability of the aircraft or the ability of the crew to cope with adverse operating conditions to the extent that there would be: 1. A large reduction in safety margins or functional capabilities, 2. Physical distress or higher workload such that the flight crew could not be relied upon to perform their tasks accurately or completely, or 3. Adverse effects on occupants including serious injury or potential fatal injuries to a small number of occupants	B
Major	Failure conditions that would reduce the capability of the aircraft or the capability of the crew to cope with adverse operating conditions to the extent that there would be, for example: 1. A significant reduction in safety margins or functional capabilities, 2. A significant increase in crew workload or in conditions impairing crew efficiency, or 3. Discomfort to occupants, possibly including injuries	C
Minor	Failure conditions that would not significantly reduce aircraft safety and that would involve crew actions that are well within their capabilities. For example: 1. A slight reduction in safety margins or functional capabilities 2. A slight increase in crew workload, such as routine flight plan changes, or 3. Some inconvenience to occupants	D

 

[Bell206]

But then I don't get to put a "DO-178C Level B" sticker on that box, which means I can't use it in "real" airplanes.

Even with a "DO-178C Level B" sticker on that box, you still couldn’t use it on a “real” aircraft without additional approvals. Same with a TSO. And technically there are no “experimental” parts. That term and other methods are simply used by venders to stay in compliance with Part 3 and Part 21. Nothing more. There is no rule that prevents you from installing a non-certified part on a TC aircraft. It just requires using a different route to install it legally.

 

[Jim_R]

Why cant avionics work as robustly as iDevices, or Dell desktops? Very frustrating.

Do a search for "iPad overheating" on pilot forums. You'll find far more posts about that than about G5s failing.

Perception bias is a thing.

And as others have explained, certification doesn't mean "guaranteed not to fail". Dud devices will make it through manufacturing quality control. Installers aren't perfect. GA aircraft present a pretty harsh operating environment.

I work with spacecraft that are built to the highest standards with jillions of dollars spent on testing. Stuff still breaks. Entropy is the law of the universe.

Also, just think about the size of Apple and the sheer number of people that are involved in the process of putting a device in your hand that "just works". Now think about the comparative size of a GA avionics company (or the avionics division of a broader company).

 

[AV8R_87]

Even with a "DO-178C Level B" sticker on that box, you still couldn’t use it on a “real” aircraft without additional approvals.

Correct. But without that sticker on your part, good luck getting an installation approval for anything else than day VFR, if that.

 

[MRC01]

I wondered the same when the old Narco DME-890 that was in my airplane when I bought it 15 years ago, started failing, getting incrementally worse reception with every flight. So I found a used plug-in replacement on eBay, which worked for about a year before it succumbed to the same behavior. Then I found yet another replacement, same thing.

Also, the MX-385 radio that I had factory restored about 8 years ago, recently started wonking out when cold, works normally about 5-10 minutes after being turned on. <sigh>

Meanwhile, all along the OEM radios in my 20+ year old cars still worked as well as they did when new, and still do.

IME, certified aviation equipment only means it has met some specification. This does not necessarily mean it is more reliable than normal consumer grade electronics. It may or may not be, and often is not.

 

[Gary Ward]

I recently had to re solder the display on my 2014 F250 radio so it was bright again. Well known problem with ford radios, the display dims out. Resolder some pins on the display and they are bright as new.
So todays car radios don't last long in this case.

 

[Bell206]

But without that sticker on your part, good luck getting an installation approval for anything else than day VFR, if that.

Not at all. We’ve been installing non-TSO equipment in TC aircraft for years. The standards you mention along with a TSO and other similar standards only apply to the production of the equipment(Part 21).

Once that vendor produced equipment is on the shelf, certified or non-certified, one simply uses an appropriate method to have that equipment installed on an aircraft legally (Part 43). Now whether it is installed as a minor alteration, or a major alteration will depend on the specific aircraft requirements to include IFR capabilities.

For example, several years ago the non-TSO G3X was certified for installation in TC aircraft via an AML-STC which included the abilities to operate IFR and couple to an autopilot if so equipped. So hardly does one need luck to get these installation approvals.

 

[hindsight2020]

lulz, I'll take that over "Never-Slide-In-Garm-In" and the gatekeeping guild remoras that feed off that sclerosis. Every day of the week and twice on AOG Sunday.

I think auto retrofit is the future of rec aviation for us sub-100K poors in this hobby anymore. This level of the aviation space will never be safe, but it's not the technical/physics moonshot being speciously characterized as [in order to justify the exclusionary expense], either.