October 13, 2010 at 7:08 am
Thanks for the article, especially since I am flying to the Bahamas to have a get away with my wife free from kids. At least we will die together!
October 13, 2010 at 7:20 am
I work for a fairly large midwestern city and have had my software featured on three separate local television news reports (as promotional stories, not attacks), and I can tell you with 100% certainty that journalists know nothing about technology. Every story contained fundamental misrepresentations about the products that had my colleagues and I rolling on the floors laughing. And this was after being interviewed by the journalists (for over a half-hour for each story).
For instance, one product allows Code Compliance officers to view aerial photos of a property with property lines overlaid to determine the responsible party of a code violation. Our aerial imagery is updated every two years. According to the story that aired our Code officers were using real-time satellite imagery to examine a property for code violations (apparently we're not big on personal privacy either).
Even unbiased journalists are required to produce a compelling narrative in a very short period of time (the stories aired the same day as the interviews), and often don't take the time to follow-up and fact-check their stories before they're published.
So even eliminating all other issues with the el Pais story, there's still a very good chance the facts are just plain wrong.
October 13, 2010 at 7:44 am
Back before seed-bearing plants, when people were actually adding NetWare 3.X servers, there were internet appliances that acted as IPX-IP gateways. We had one.
Some people said, "But you don't have a real connection! Think of the integration you're missing!"
I replied to the effect of, that's not a bug, that's a feature. We did not worry about remote access to internal hosts. They couldn't hear the script kiddies at all.
October 13, 2010 at 7:45 am
Agreed that often the media gets things incredibly wrong without doing simple back of the envelope calculation to see if it even makes sense.
Like the article a while back claiming that traffic violators would be spotted by satellite
But there is a point here that the extensive use of computers has introduced vulnerabilities and points of failure that did not exist before. This is an area of real concern. There is a push to have more and more auto systems, for example, removed from actual control by the driver. Points of failure increase, and while computers are fairly reliable, when they fail (or when algorithms fail) the results can be spectacular. (one example are the 'vehicle stability' systems, which use inputs from acceleration sensors and wheel sensors to decide to take control away from the driver. Essentially a blind system, who does not see what you see, deciding what is happening).
Another area that has been examined in aircraft is the potential for human failure increasing with more automated systems. Boeng for example has actually NOT automated everything possible because it causes inattention on the part of the flight crew. Keeping that in mind, I am concerned that these automobile radar based 'accident avoidance' systems may actually lead to inattentive driving. (And they are far from foolproof. TWO recent demos, by Volvo, in front of reporters, failed dramitcally. Fortunately since it was a test track, no one was injured.)
...
-- FORTRAN manual for Xerox Computers --
October 13, 2010 at 7:47 am
The article was definitely sensationalistic. However I linked to Schneier's blog since he has some better articles in there as well that talk about pilot issues with this crash.
My point wasn't that this article was correct and software crashed the plane, but we ought to be aware that this could be an issue. Not is, but could. As we build more software into our mechanical transport systems.
October 13, 2010 at 7:52 am
Wish I had read this one AFTER safely returning from my trip next week. 🙂
October 13, 2010 at 8:59 am
Steve, you could do like I did (after I met you at the Rockies game), and take the train! 😀
Seriously, I understand your point; systems need to be secure, lest it become another point of failure. FWIW, I've seen stories about entire radar systems failing, with the FAA responding with something like "passenger safety was never compromised," etc.
I will say that, although critical computer system security is a legitimate concern, it would hardly be, by itself, the cause of a catastrophe. My understanding is that a major accident, such as a plane crash, is never attributed to one single event, but usually a perfect storm or chain reaction of multiple events. Even if a system goes down, how the crew reacts to it can make the difference between a 737 landing safely and flying into a mountain.
+--------------------------------------------------------------------------------------+
Check out my blog at https://pianorayk.wordpress.com/
October 13, 2010 at 9:08 am
There are aircraft, mostly fighters, for which failure of the flight control computers in flight would result in catastrophic airframe failure. These are triply redundant computers, but if all three DID fail, you could not even glide: the plane would immediately depart controlled flight in such a way that the wings would come off (forward sweep).
October 13, 2010 at 9:15 am
edward.ahlsen-girard.ctr (10/13/2010)
There are aircraft, mostly fighters, for which failure of the flight control computers in flight would result in catastrophic airframe failure. These are triply redundant computers, but if all three DID fail, you could not even glide: the plane would immediately depart controlled flight in such a way that the wings would come off (forward sweep).
Actually even in commercial aircraft there is a similar problem. The Air France crash was caused initially by a failed airspeed indicator, and as I understand it the computer allows very little override. Sort of ok, untill you get to where the computer is responding to its perception of the world. Mix that with a virus (even a relatively non malevolent one) and you can really have problems.
...
-- FORTRAN manual for Xerox Computers --
October 13, 2010 at 9:30 am
And the latest car ads on TV promise to build cars that will avoid accidents. Wait until these babies hit the road, then you will hear all sorts of stories about malware accidents.
October 13, 2010 at 9:42 am
ron.carlton 41182 (10/13/2010)
And the latest car ads on TV promise to build cars that will avoid accidents. Wait until these babies hit the road, then you will hear all sorts of stories about malware accidents.
http://jalopnik.com/5648126/volvo-pedestrian-avoidance-crash-test-fails-spectacularly
Funny only because it was a demo on a test track.
...
-- FORTRAN manual for Xerox Computers --
October 13, 2010 at 2:15 pm
And yet flying is likely to still be the safest part of your trip 🙂
There are a couple of things that bug me about the integration of computers and things like cars and planes. The first one might sound a little bizarre: I know quite a few people that I would trust to build a roadworthy car and even one or two that I would trust to build am airworthy plane, but I don't know anyone, including me, that I would trust to build a computer-assisted control system for either of those things. (Note that I don't consider carefully engineered crumple zones and airbags to be part of 'reasonably safe' even though I make purchasing decisions based on crash-test ratings.)
The second is a bit more mundane, but it goes a long way to fostering distrust of such systems because it shows that even the basics are badly done. My car (2009 Subaru Forester w/ STD transmission) has a drive-by-wire throttle. There is so much lag in the response to pedal position that I had to relearn how to shift so that I was properly matching rpm to the newly selected gear.
October 13, 2010 at 3:46 pm
So long as we can stay away from the sensationalist approach to all of this and hold a rational discussion then I'm fine.
But this does put me in mind of a recent trip to the car mechanics to have a factory recall on my wife's car corrected (10 minute job, firmware upgrade) and I was shocked to see the car standing off to one side with no-one near it. I asked the service manager when they would start work, and he told me they were nearly finished. They had plugged a bluetooth dongle into the car and were updating the firmware from a laptop on the other side of the floor.
A far cry from the mechanics of my youth who were covered in oil.
October 13, 2010 at 4:55 pm
An interesting and informative set of posts.
My initial reaction was triggered by the fear that too much opinion is formed based on repetition of half-baked conclusions from downright misleading stories.
The ungrammatical use of 'less' rather than 'fewer' reinforced the idea of a piece that hadn't been thought through.
Viewing 15 posts - 16 through 30 (of 36 total)
You must be logged in to reply to this topic. Login to reply