Airstream interstate - is it worth the hype?

[avanti]

Originally Posted by hepcat

Unfortunately, with electronics driven by computer systems, you're at the mercy of electrical connections, chip, sensor, and circuit quality control, and the foibles of human programmers. When that equipment and those instructions apply only to the operation of the engine and transmission, a failure means that the vehicle stops running. That can be really bad, depending on where and when it happens, but isn't on the same magnitude of the car's systems giving the steering a nudge, or braking on its own, or taking other autonomous action without the knowledge, consent, or input of the driver. And while there may be a programming team in place for a specific manufacturer, we have NO idea how competent or experienced they are... and like a piece of chain, the software is only as competent as it's weakest piece of code.

[...]

I guess my objections are as much on principle as an active mistrust of electronics. I do NOT want my car to try to correct my driving for me. I DO expect the other drivers on the road to take their job of driving the car seriously. Unfortunately, I know that too many don't. If they did, we wouldn't need to be paying for active driving intervention safety systems that rely on inherently unreliable technology.

I guess we are going to have to agree to disagree on this one. The idea that electronics are "inherently unreliable" is just nonsense. We routinely build spacecraft with electronics that consistently operate in incredibly hostile environments for many decades with zero PM and no possibility of repairs. The laptop I am typing on has roughly 200 billion transistors and hundreds of mechanical connectors. The failure of any one of them is likely to render the device useless. At the component level, the MTBF is astronomically high. If we can do THAT, there is absolutely no reason to think that we can't learn to make ANY arbitrarily complex electronic system arbitrarily reliable at an arbitrarily low price. If there were some kind of "wall in the sky" to prevent this, we would have found it by now. It is just a matter of time and experience.

For every "electrical connection" in an electrical system, there is a "fastener" in the analogous mechanical system. "Chip, sensor, and circuit quality control" issues all have similar analogs. And, "the foibles of human programmers" are just a special case of the more general issue of the "foibles of engineers." All those crumple-zones that you like so much are just as dependent on correct software as anything built out of transistors.

There is simply no inherent difference in reliability between electrical and mechanical systems. There are only differences in product maturity.

If you are inclined to throw your wooden shoes into the gears, that is your choice. But once you buy into the whole technology thing, to believe that one wave of technology is "inherently" more reliable than the next is simply incorrect.

As an aside, I am curious why you would not want a car to correct your driving, stipulating that it was better at it than you are. You may disagree with the stipulation, but it seems clear that you would be on increasingly thin ice. When a self-driving Tesla has ONE SINGLE fatality it makes headlines. In the mean time, the slaughter caused by human drivers continues without notice. If you examine the accidents-per-mile-driven statistics, you will see that these technologies are ALREADY far more capable than human drivers. It will soon become no contest.

[booster]

It is an interesting comparison of the electronics vs the mechanicals. I spent nearly 50 years in manufacturing plants and was always involved in sourcing, maintaining, and repairing the equipment. Back then, the fanciest electronics you saw were switches and relays, so a it was a lot simpler time.

If you look at it from a capability standpoint, there is no comparison. The newer PLC and other computer controls and sensor technology have totally revolutionized manufacturing and the associated equipment. I think it is the same with vehicles, although not as extreme.

If you look at it from a reliability standpoint, a lot depends on how you define reliability. It the recent years, how mechanical systems were handled changed some, particularly from predictive maintenance and monitoring systems, but what didn't change was that you often, but not always, had a pretty good idea of what was going to fail or deteriorate. They mechanical systems could also mostly be repaired by in house maintenance and tech employees.

With the electronic systems, life was a lot tougher for us that had to keep the production output happening. Integrated, electronics controlled, equipment is much harder to keep at the very high uptime targets the bosses want and need to compete. There are several reasons, but mainly that you tend to not get as much warning of impending failures in most cases, and on site personnel are usually not capable of repairing the systems requiring bringing in high end contractors or factory techs from the manufacturers. The delays in getting that kind of service can easily blow you uptime percent to a whole quarter just in one failure. The failures are usually also much more expensive to repair.

So, IMO, even if the electronic and mechanical failures happen at similar frequencies, which is probably pretty accurate, we found the electronic ones to make our lives much worse in the factories.

I have no idea how many of the fancy electronic things now being used (both for safety or giltz) fail as a percentage, by year. That would be interesting to know, I think. I do know that when they do fail, it often is very expensive to get them fixed, with electronic repairs getting past the cost of engine and transmission rebuilds in some cases.

You can't put a price on safety, especially if you are the one saved, but safety doesn't come cheap.

A question that will interesting to see the answer to is if all the new systems will make the drivers less attentive and more careless, negating some of the gains from the features. The Tesla guy that died because his Tesla didn't see a semi trailer against the white sky, comes to mind on that issue, as he was allegedly driving and he missed it too. How do you miss a big semi and trailer in broad daylight unless you are not paying attention at all.

[avanti]

Originally Posted by booster

A question that will interesting to see the answer to is if all the new systems will make the drivers less attentive and more careless, negating some of the gains from the features. The Tesla guy that died because his Tesla didn't see a semi trailer against the white sky, comes to mind on that issue, as he was allegedly driving and he missed it too. How do you miss a big semi and trailer in broad daylight unless you are not paying attention at all.

Yes, this is a big issue. I think it is the biggest challenge that the designers have in rolling out the current generation of "semi-autonomous" systems--keeping the driver in the loop without being annoying. The same issue arose 50 years ago in locomotives, when diesels got simple autopilot systems. Engineers promptly went to sleep. They had to add various "dead man switches" and "I'm still here" response systems.

As I noted before, however, sometimes the results are virtuous. I find that lane-keeping has actually IMPROVED my attention by giving me good feedback as to the results of my inattention.

Of course, in the long run, none of this will matter--the vehicles will not have steering wheels.

[hepcat]

Originally Posted by avanti

I guess we are going to have to agree to disagree on this one. The idea that electronics are "inherently unreliable" is just nonsense.

"Chip, sensor, and circuit quality control" issues all have similar analogs. And, "the foibles of human programmers" are just a special case of the more general issue of the "foibles of engineers." All those crumple-zones that you like so much are just as dependent on correct software as anything built out of transistors.

There is simply no inherent difference in reliability between electrical and mechanical systems. There are only differences in product maturity.

I guess you're right. Reliability in electronics is reported in "MTBF" or Mean Time Between Failures. As the technology matures, the MTBF gets longer... but each component still has a failure rate. If that weren't the case, there'd be no reason for auto makers to stock replacement units, or for TV manufacturers to offer warranties.

I just don't want to be on the -3 end of the MTBF standard deviation curve in a car that I'm driving.

Crumple zones and other such passive systems are fixed. They don't turn off and on... there's no MTBF rate. They're designed and tested... and the design either works or it doesn't.

Originally Posted by avanti

If you are inclined to throw your wooden shoes into the gears, that is your choice. But once you buy into the whole technology thing, to believe that one wave of technology is "inherently" more reliable than the next is simply incorrect.

I'm sorry, but I just can't agree with this. Metallurgy, machining, and engineers skilled in their craft have been around for a couple of hundred years. Computer technology has only been around in a society-changing way since about 1990. I remember being in a big-box electronics store in Sacramento in about '94 or so when a car crashed into a transformer that fed electricity to the store and killed the power. The whole store went dark... and the emergency perimeter lighting came on. I realized at that point how useless electronics are without a power source.

Originally Posted by avanti

As an aside, I am curious why you would not want a car to correct your driving, stipulating that it was better at it than you are. You may disagree with the stipulation, but it seems clear that you would be on increasingly thin ice. When a self-driving Tesla has ONE SINGLE fatality it makes headlines. In the mean time, the slaughter caused by human drivers continues without notice. If you examine the accidents-per-mile-driven statistics, you will see that these technologies are ALREADY far more capable than human drivers. It will soon become no contest.

Well, I think you hit the nail on the head. I don't have the confidence in the technology that you obviously have. And the Tesla crash made headlines precisely BECAUSE it was a bias-confirming incident for many of us who aren't yet ready to trust the programmers just because the salesmen tell me that they can be trusted.

iOS 10.3 had more beta testers and beta testing than ANY previous iOS version released after the programming disaster that 10.2.1 had become, and yet, within a week Apple released 10.3.1 to fix bugs in 10.3. I don't trust my physical safety to my phone... I DO trust it to the manufacturers of my car... and we're talking about essentially the same technology in my phone put to work doing different tasks in the automotive industry.

So, yes, it IS still inherently unreliable technology and yet we've now based the majority of our existence on it.

[avanti]

This is my last comment on my topic--you are welcome to the last word...

Originally Posted by hepcat

So, yes, it IS still inherently unreliable technology and yet we've now based the majority of our existence on it.

This is simply crazy talk. By your standard, ALL technology is "inherently unreliable". If you think that your crumple-zone is "fixed" and that it either "works or it doesn't", you are deluded. First of all, it will respond differently in every situation, and will work better or worse depending on a thousand factors. More to the point, components wear, they are subject to metal fatigue, they rust, etc. etc. They most certainly DO have MTBF values. THERE IS NO DIFFERENCE between one technology and another. What, exactly, do you think defines the difference? The use of electrons? The chemical silicon? or what? It is all just physics. There is no reason to worry about being at -3 standard deviations, if the mean is high enough.

All of the examples you give reflect not "inherent" differences, they reflect differences in maturity. I have already said that I think it is fine for you to decide to live at some distance behind the bleeding edge. I do not judge. But don't blame the electrons when it is just history.

Over and out.

[GeorgeRa]

Sorry I am taking some liberty to continue on this subject but will bail out after this post.

Practically each technology has the beginning and the end and folks react differently during different period of technology lifecycle bell curve. Some of you are likely familiar with “Crossing the Chasm” by Jeffrey Moore. See this graph - Technology Adoption Process | Applitude

Just count how many music recording technologies has passed during the last 150 years. From metal discs with punched hooks to a semiconductor memory.

I am reasonably familiar with Ink Jet printing technologies, one of the most complex devices in electronic industries are now ubiquitous. I remember the first TI digital images based on array of moving mirrors with each one representing one pixel, now they are in cinemas.

George.

[avanti]

Originally Posted by GeorgeRa

Heh. I own one of these:





Anderson Jacobson Model ADC-300 modem - CHM Revolution

[BBQ]

.

I hate to admit it,

I have used a 300 baud acoustic coupler modem.

I used to have one at home driving a Teletype terminal and my 8080 computer. LOL

[booster]

We have already discussed reliability and safety, but those won't be the biggest problem with the car electronics and features that some of use will have.

We drive cars until dead, and 20 years is not at all out the question. It can get very hard to find the proprietary electronics after a number of years, and if it is a non big seller vehicle, the electronics become unavailable sooner and extremely hard to get later, especially if they are trouble prone. If it is a feature that disables the vehicle or makes it unsafe to drive, you are basically all done if you can't get the parts.

[InterBlog]

Long thread... has anyone noted that all of the technology being debated somewhat amorphously here is not, in fact, created equally?

Because it's not a question of "is technology inherently unreliable" or is it not. In order to make a rational, informed decision about whether any given technology is worth relying on, you first have to know what that tech is, and whether the team that installed it knew what they were doing. And the answer to that second question in many cases is "no". There's a shockingly large range in quality in what we've found installed in our Airstream Interstate, getting back to the OP's original question.

Here is an example, with photograph, of a thread entry where my husband refers to an Airstream-installed technical component as a POS. It took us months to discover and locate the source of that particular electronic failure. It would not matter how well-designed or reliable the other technical components were, that were installed in line with that one. That was the weak link in that particular chain.

Tech is either reliable or unreliable depending on (1) how it was designed, (2) how it was constructed, and (3) how it was installed. It's not possible to generate blanket reliability conclusions one way or the other in the absence of understanding those situationally-specific factors.

[hepcat]

Originally Posted by InterBlog

Long thread... has anyone noted that all of the technology being debated somewhat amorphously here is not, in fact, created equally?

Because it's not a question of "is technology inherently unreliable" or is it not. In order to make a rational, informed decision about whether any given technology is worth relying on, you first have to know what that tech is, and whether the team that installed it knew what they were doing. And the answer to that second question in many cases is "no".

And you've really hit on the crux of the issue... faulty installation is JUST as fatal a flaw as faulty design, faulty materials, or faulty manufacture. And unfortunately, the consumer often can't distinguish among them... nor should they have to.

From a consumer's perspective, if the part doesn't work when it's employed, it's unreliable and the cause really doesn't matter. And unfortunately these systems are SO complex, the miracle isn't that there are relatively few failures, but the miracle is that they work at all.

I stand by my statement that computer technology remains inherently unreliable; partially because there are SO many entry points for failure, including installation. And despite the fact that there are a lot of installed units of all kinds working fine in all manner of service today, I don't think you'll find anyone who uses this technology, either actively or passively, who hasn't experienced some kind of failure of computer technology; most of us many times over.

[InterBlog]

It's also worth considering that none of us have access to statistics that would allow generalized conclusions about overall reliability, as that reliability is influenced by multiple factors including installation.

One of the newest threads on Air Forums is titled "Blind spot assist on 2016 [Airstream Interstate] Grand Tour not working". Splendid - Airstream (edit: MB) installed this fancy safety tech and less than two years later, it's dead on that particular rig. But we don't know how many other rigs will experience the same outcome. Certainly that's not the first reported case of that particular failure (I've seen previous posts regarding it), but how many are there, actually? We simply don't know. If it's one failure per hundred units, maybe I'd personally want to take a chance on paying for it. But if it's one failure in ten, probably I would not.

Some of both the actual and the potential failure pathways we've noted during our extensive Airstream Interstate DIY... we would estimate those as being higher than one failure in ten to be expected. That's a real problem.

[hepcat]

Originally Posted by InterBlog

Some of both the actual and the potential failure pathways we've noted during our extensive Airstream Interstate DIY... we would estimate those as being higher than one failure in ten to be expected. That's a real problem.

The MTBF rate and the rate of failure among installed units is a logistics problem, but for we drivers, the REAL problem with this technology in general is that it works frequently enough that we learn to rely on it... gps units and ABS braking are great examples... and then when it fails at a critical moment when we're relying on it, the user may not be able to recover from the failure.

Thankfully most applications of technology aren't life-critical; ATMs for example... but in MY van, something in the ABS is wonky and the ABS cuts out at the most unpredictable times. It doesn't store a permanent code in the OBDI system though, so it's more difficult to diagnose. Likely a wheel sensor... but fortunately I don't rely on "stomp and steer" anyway, so it's not critical for me... but if that's how you learned to drive, and you base your driving habits on "stomp and steer" expecting the ABS to modulate your braking... and it doesn't... that could be a problem.

And that's the problem with lane-keep-assist, radar cruise, and some of those other delightful technologies... when you let your mind drift off to la la land knowing that the car will take care of you because it's trained you that it'll take care of you, and it doesn't... all of a sudden you've got real problems.

[Taratupa]

I wonder if tympa, the original poster, hung himself in lieu of buying any RV, after seeing the way this discussion almost immediately devolved wildly from his posted question.

[tympa]

Actually this chain has been really interesting. I think that technology is great but it isn't perfect and it can fail like everything else including Human error. I still prefer a RV with safety features but just wonder about the cost benefit ratio.
I am considering the Airstream despite its quality concerns because I like the safety features and I couldn't find another RV that comes with the safety features other than a fully custom Rv build. Is there another company that has the safety features? Thanks again for all your comments!

[cruising7388]

Originally Posted by tympa

Actually this chain has been really interesting. I think that technology is great but it isn't perfect and it can fail like everything else including Human error. I still prefer a RV with safety features but just wonder about the cost benefit ratio.
I am considering the Airstream despite its quality concerns because I like the safety features and I couldn't find another RV that comes with the safety features other than a fully custom Rv build. Is there another company that has the safety features? Thanks again for all your comments!

I think any Roadtrek on a Sprinter platform has the advanced safety features and if you're not committed to Etrek technology, a coach worth taking a look at is the safety equipped Avion Azur.

[avanti]

Originally Posted by tympa

Is there another company that has the safety features?

Pleasure way Plateau

If I were buying a b-van today, the Plateau is probably what I would choose (assuming I couldn't find an unsold GWV Legend).

[hepcat]

avanti, I too have over-posted on this thread... but I have to make one more.

Here's another bias-confirming case in point:

Fiat Chrysler Recalls 1.2 Million Ram Pickups Over a Fatal Flaw

Tympa, I think your rationale is sound. I appreciate the technology as well, and the ideas behind it have merit but I don't think it's mature enough yet to rely on which diminishes the cost/benefit ratio dramatically for me.