Wikipedia:Reference desk/Archives/Science/2013 May 25

Science desk
< May 24	<< Apr | May | Jun >>	May 26 >

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

May 25

Are super-tankers fully automated ?

Hi Everyone.
In Superman III and Hackers (and many other fictions, I guess), it is said that modern oil tankers can not be controlled manualy and are totaly driven by computers. I wonder if this is accurate. I know that in the 1970s there have been numerous tankers hijacking stories and I guess automation is a way to avoid that ? If there is a super-tanker captain here, I'd be glad to here what he says about that ;-) Jean-no (talk) 13:43, 25 May 2013 (UTC)

If by fully automated you mean so automated that computers set the course in a way that humans can't override, certainly not. That would be insane. See Exxon Valdez oil spill, among other things. Looie496 (talk) 14:40, 25 May 2013 (UTC)

According to the article on that spill, the main causes were human fatigue and failure to maintain equipment. It seems like full automation may have mitigated or even prevented it. 202.155.85.18 (talk) 02:18, 27 May 2013 (UTC)

In the movies I mention, the ship's course is decided by distant and centralized computers. It seems quite dangerous to me, so I guess the writers didn't understand well something... Jean-no (talk) 15:01, 25 May 2013 (UTC)

I suppose by automated they could mean computerized, such that the controls could be hijacked. So not like the captain and crew does nothing, but more like Fly-by-wire for boats. Mingmingla (talk) 15:22, 25 May 2013 (UTC)

There are (at least) two separate questions there. The first is whether or not such functions are computerized/automated, the second is whether or not the computer is required to carry out those functions: adjusting ballast, controlling the engine and rudder settings. (The third implicit question, I suppose, is whether or not those functions are, or can be, remotely controlled. My common-sense response is "Why the hell would you put that on the internet?", though dumber things have happened. Absent any evidence to the contrary, I would be very reluctant to presume that the capability exists; there probably isn't an idiot oil-industry exec who has said "Please, can I adjust ballast on a supertanker from my iPhone?") TenOfAllTrades(talk) 16:40, 25 May 2013 (UTC)

My unschooled thought is that more security would be needed than you'd think. At some point the computer now on board the tanker could have been connected to the Internet, and somebody might have had a pop-up "Would you like to update your map of the reef to include the most recent soundings?" and not realized that someone was spoofing the update site and uploading a map that shows clear water sailing straight through Diamond Head. The ship probably has an autopilot, which could remain engaged if the crew had ... an accident. Indeed, the 'accident' might well have been loaded on board together with the new navigation computer. I suspect the movies were more dramatic than this, though! Wnt (talk) 17:47, 25 May 2013 (UTC)

Movie plot threats aside, do you actually know anything about how navigational or control systems on board a supertanker work? TenOfAllTrades(talk) 22:39, 25 May 2013 (UTC)

I didn't know the expression "Movie plot threat", I love it. Jean-no (talk) 22:53, 25 May 2013 (UTC)

The real movie plot threat is the Terminator 2: Judgement Day! The singularity is near!Shadowjams (talk) 00:53, 26 May 2013 (UTC)

To be clear - no. I just wanted to keep people's minds open about the idea in theory, as SteveBaker has done at more length below. Wnt (talk) 18:44, 26 May 2013 (UTC)

Several approaches exist to infecting and controlling computers that are not, nor have ever been connected to any networks whatever. You can even devise attacks for computers that are physically incapable of being connected to any outside communications whatever!

This seems unlikely - but it's actually happened. The most famous is the Stuxnet system which was an effort (allegedly by the US and/or Israel) to disrupt uranium enrichment plants in Iran that were controlled only by very simple microcontroller chips with no network capability whatever.

It worked by first infecting computers that WERE connected to the Internet, then having those machines place malware on things like thumb-drives, cellphones and MP3 players that were plugged into those machines - then when someone who designed automation software for Iran charged their infected phone or MP3 player by plugging it into their (network-isolated) software design workstation, the infection was spread into the software that was used to program the dedicated microcontrollers - which in turn caused covert software to be placed into their firmware that caused them to fail under some future circumstances and to hide that failure from instrumentation designed to see faults exactly like that!

That approach requires massive amounts of patience and investment in time and money to get all of those steps to work correctly - but the outcome in that case is well known to have been moderately successful.

It seems entirely reasonable to me that a similar trick - planned years in advance by some kind of James Bond super-villian - could allow one to give a super-tanker some predefined commands that would be hard for the crew to diagnose or rectify without knowing what was happening and why.

That said, the emphasis would be on "predefined". It would likely be impossible for any degree of control over the tanker to be had remotely during such an event. But a sufficiently cunning bad guy could certainly have the thing subtly veer off course and crash into something important unless the captain enters some special code-word into the computer via whatever controls he has. A system like Stuxnet is capable of doing just that. Now, obviously if it took control of the tanker hours before this impact, the crew could disconnect the computers and shut down the engines by cutting power lines with an axe or something - but if it jammed the bow thrusters into full power right in the middle of threading a course through some narrow gap between two rocky outcrops in the middle of the night - and the radar and other navigation computers were rigged to make it look like the correct course were being followed - it could probably cause the needed mayhem before the crew could figure out what was happening.

The plot of that Bond movie where the British warship is sent off-course into Chinese waters by doing something nasty to its GPS is exactly the kind of thing that a Stuxnet type of system could use to have a supertanker run aground and cause an ecological and political nightmare someplace.

Basically, it *is* technologically plausible...but unlikely.

SteveBaker (talk) 02:56, 26 May 2013 (UTC)

Again I ask, does anyone know anything specific about the degree of automation or computerization of the controls of a supertanker or large freighter? Is there actually a USB port on the bridge where a crewmember can plug in his mp3 player and run arbitrary, malicious Windows code on the navigational and control systems simultaneously? I can't speak to the plausibility of Steve's scenario, because – correct me if I am mistaken – no one here has actually done any research into the questions asked, nor admitted to so much as working aboard a ship, let alone ever reflashing a supertanker's firmware. TenOfAllTrades(talk) 19:14, 26 May 2013 (UTC)

The lesson we should take away from Stuxnet is that even if the supertanker's computers had no internet or USB ports, they could still have malicious code placed there when the ship was first built if the computer on which the supertanker's software was written was connected to the internet or had an infected device plugged into one of the USB ports. However, the skills and resources needed to make that happen are huge - and the need to plan all of this years in advance - make it quite unlikely that this kind of exploit has ever actually been used against supertankers. But it's definitely possible. SteveBaker (talk) 12:52, 28 May 2013 (UTC)

Well, it's something of a hypothetical question -- pirate attacks do occur, but they actually do encounter a crew. (in one case, 25) Of course, the way pirate attacks usually go, the crew ends up being a liability as hostages, not really a guarantee nothing bad is going to happen. Wnt (talk) 21:13, 26 May 2013 (UTC)

For an interesting parallel, see [1]. That's about airplanes, but the point is, apparently in the corporate world nobody really cares what could happen, as long as they can get through tomorrow. Wnt (talk) 03:56, 27 May 2013 (UTC)

I just love it when folk include links to articles as though the articles support the nonsense they claim. In this case there is a slight twist: The news item cited by Wnt does state that someone claimed you can misdirect an aircraft with a phone. But there is a link at the bottom of it to another news item issued a couple of days later. It quotes an official American FAA press release and the relevant systems manufacturer - both state that its a load of codswallop - you simply cannot take over a plane with an andriod phone - and some technical detail why is given. Wickwack 120.145.154.162 (talk) 10:59, 27 May 2013 (UTC)

Thanks for pointing out that response, but it isn't as firm a denunciation as you think. "Neither the FMS nor the autopilot flies the plane. The crew flies the plane through these components. We tell it what to do, when to do it, and how to do it. Whatever data finds its way into the FMS, and regardless of where it's coming from, it still needs to make sense to the crew. If it doesn't, we're not going to allow the plane, or ourselves, to follow it." No, the plane won't do loop-the-loop if you drop your accelerometer-enabled Android phone in the toilet. But how about if the weather is foggy and the crew thinks they're landing on a runway that is just barely long enough for the plane but actually it's another runway at that airport that isn't long enough, because they're not quite where the instruments say they are? Wnt (talk) 13:48, 27 May 2013 (UTC) (I suppose I shouldn't leave without a mention of Die Hard 2...)

Yes, it is a firm denuciation. The perpetrator said there is NO security on these systems. But the FAA and the manufactuer says there is. I'll believe the FAA before I rely on some nutter who gets the ear of a journo. I can think of multiple reasons for not believing this rubbish anyway. Here's the first one: How is the andriod phone hardware to communicate with the airplane systems? They do not share frequencies. That does not preclude some villian with a purchased or custom-made system that IS on the right frequency of course, but that requires a lot more resources, and technically sophisticated ones at that. Wickwack 120.145.148.85 (talk) 15:02, 27 May 2013 (UTC)

I don't know how myself (but half the time "apps" discussed in this manner end up involving some kind of hardware accessory, e.g. for terahertz imaging). However, I leave it to readers of the article [2] to decide themselves what it says. I think it is pretty clearly saying that access to the FMS does not allow direct control of the autopilot or plane, not that it can't be hacked into or mislead pilots. Wnt (talk) 17:07, 28 May 2013 (UTC)

Well, as I said before, but apparently you missed it, it says very clearly in the 7th paragraph that aircraft systems have high levels of security. It was the perp's claim that there was no security. So even with the right transmission hardware, his claims are simply not true. Wickwack 124.178.175.45 (talk) 00:08, 29 May 2013 (UTC)

I've just spoken to a youth who is studying to be a Merchant Navy officer, and has been to sea a couple of times in bulk carriers. He says that the navigation system is along similar lines to an aircraft autopilot and that in some ships, the control functions are computerised, in the manner of fly-by-wire in aircraft. He doesn't think it would be possible to hack into a ship's systems. Alansplodge (talk) 14:10, 28 May 2013 (UTC)

• Followup: http://www.foxnews.com/tech/2013/07/26/exclusive-gps-flaw-could-let-terrorists-hijack-ships-planes/. Wnt (talk) 19:49, 28 July 2013 (UTC)

Solubility of Iron(II) oxalate

Iron(II) oxalate says it is slightly soluble, but Oxalate#Physiological effects says highly insoluble. Which is correct? — Preceding unsigned comment added by 93.132.188.84 (talk) 15:47, 25 May 2013 (UTC)

It's plausible that both are correct, insofar as they are subjective, context-specific descriptions. Iron oxalate obviously has a specific solubility constant in water; but it's not available in NIST's online database. The next step would be to check a handbook of chemistry and physics constants (e.g., the CRC Handbook of Chemistry and Physics); maybe one of our chemistry-minded regulars has a copy of such a resource handy. Nimur (talk) 16:23, 25 May 2013 (UTC)

Here it is at PubChem: FERROUS OXALATE, again only with subjective description of the solubility. Nimur (talk) 16:33, 25 May 2013 (UTC)

It's a bit of a shock if both should be correct in the respective (and highly unspecified) contexts. 93.132.188.84 (talk) 16:52, 25 May 2013 (UTC)

What's the difference between slightly sober and highly drunk? Seriously, words like "slightly" and "very" are quite imprecise, and there's nothing that makes "slightly soluble" and "highly insoluble" incompatible. If your really want to know, you need the K_sp value for Iron (II) oxalate. This says the K_sp value is 2 x 10^-7. Since this is a 1:1 ionic compound, that means that the molar solubility is SQRT (K_sp), which would be 2 x 10^-7/2 = 6 x 10^-4 mol/L. --Jayron32 18:39, 25 May 2013 (UTC)

So wouldn't it be better to avoid those imprecise (weasel-) words in those articles? At least in those fact-boxes for chemical substances? 93.132.188.84 (talk) 18:52, 25 May 2013 (UTC)

I'm neither here nor there on that. Many sources use terms like "slightly soluble". But if you wanted to add a bit to directly list the Ksp, and link to the reference I gave, additional information is always good. --Jayron32 19:04, 25 May 2013 (UTC)

Last few times I tried to edit anything on wikipedia failed due to one restriction or the other. I'd rather leave that to the "insiders". 93.132.188.84 (talk) 19:19, 25 May 2013 (UTC)

This reminds me of the question I asked earlier. The answer seems to be that as adjectives, 'very', 'highly', 'slightly', and the words 'soluble' and 'insoluble' are highly subjective terms. Other subjective words that also pop up frequently in science, include 'inorganic' and 'warm/hot' and 'metalloid'. Plasmic Physics (talk) 09:00, 27 May 2013 (UTC)

Universal constants are very important to effective scientific communication, but it seems that these words have escaped attention. Even arbitrarily bounded definitions would prevent such confusion. For now, they remain the bain of science, subject to the various schools of thought, and the rogue individual. Plasmic Physics (talk) 09:06, 27 May 2013 (UTC)

No, the words are fine. What you really need to know about something at first glance is what its solubility is based on three basic levels:

• If I take a spoonful of it and stir it into a big cup of water, will it essentially all dissolve
• Will it not dissolve completely, or even much at all, but will it still produce measurable or chemically significant amounts of ions in solution
• Will it do absolutely nothing, and produce no measurable or chemically significant change to the water

The first condition is always "soluble". The last condition is always "insoluble". Depending on why you are asking the question, the second condition could be "insoluble" or "slightly soluble". Context matters, but it doesn't mean that the words are completely useless. --Jayron32 18:49, 28 May 2013 (UTC)

Organ theft

Has there ever been a confirmed case where someone has been convicted of kidnapping people and stealing, then selling their organs on the black market? I've had a look around the web, and I see that people have occasionally been *accused* of it, but then the true story turns out to be something less sensational when it goes to trial.

I'm not talking about the urban legends of serial killers being paid by the Russian mafia/Triads/etc. to abduct hobos and teenage runaways from the streets of Western cities and 'harvest' their organs for sale to rich people in the far east either. I've heard that stuff before and I'm pretty sure that it is 100% 'hooey'. Or travellers who get their drinks spiked and wake up in motel room bathtubs having had their kidneys stolen. --87.112.25.55 (talk) 22:30, 25 May 2013 (UTC)

Organ theft in Kosovo. TenOfAllTrades(talk) 22:32, 25 May 2013 (UTC)

You may be interested in looking at the work of Nancy Scheper-Hughes. --Atethnekos (Discussion, Contributions) 06:50, 28 May 2013 (UTC)