I’ve gotten used to being cited in computer science and software engineering papers over the last decade, but here’s a new one. Today I read a draft in which I and the GPSD project get cited a bunch of times and it’s – er – not about open source. It’s about Marine AIS in disaster management. Broadening my deepwater horizons, as it were.
You can read a PDF of Vessel Tracking Using the Automatic Identification System (AIS) During Emergency Response: Lessons from the Deepwater Horizon Incident, but be warned that it’s pretty heavy going unless you’re deep into AIS or have a thing for disaster-management porn.
It’s nevertheless interesting on several levels. Kurt Schwehr, the author, writes well for an academic and manages to make the bureaucratically-clogged portions of the narrative less deadly boring than they might have been. Both the low-comedy aspects of government reaction to the disaster and the heroic attempts to cope with the clusterfuck-in-motion come through pretty clearly.
I’m only a supporting player in this story, but the paper does illustrate one point I’ve been hammering on for years. Making standards for life-critical systems proprietary is not just stupid, it’s evil. Kurt reports that some work I and the GPSD project guys did helped prevent that from being a problem in his part of the Deepwater Horizon response, and I’m pleased about that. Next time, though, we might not get so lucky.
Unanswered question: was GPSD running in any of the robot submarines they sent in to try to plug the gusher? Inquiring minds want to know…
Erm. I’m sure you meant to link http://vislab-ccom.unh.edu/~schwehr/papers/2011-schwehr-ushydro-dwh.pdf
(You forgot to close your anchor tag).
The link for the pdf is “http://vislab-ccom.unh.edu/~schwehr/papers/2011-schwehr-ushydro-dwh.pdf”.
I’m pretty sure GPSD was not running on sumbarines. GPS is useless under surface and the submarines are not designed to travel the seas without help of mother ship…
>I’m pretty sure GPSD was not running on sumbarines.
Try being less sure. It’s in use on the Gavia family of AUVs and the Woods Hole Nereus vehicle, the deepest-diving AUV in the world.
AUVs often use GPS for near-surface navigation and combine this with an inertial tracker for deep-water work. The above two are only the AUV deployments of GPSD that I know about; there are very likely others as well.
I can confirm that this exactly what military AUVs do as well. I can’t 100% confirm that the military uses GPSD in any AUVs, but I wouldn’t doubt it since I do know that the military uses Linux-based firmware in some of their more recent navigation systems.
I can neither confirm nor deny… ;-)
> I can neither confirm nor deny… ;-)
I could tell you.
But
thenfirst I’d have to kill you.>AUVs often use GPS for near-surface navigation and combine this with an inertial tracker for deep-water work. The above two are only the AUV deployments of GPSD that I know about; there are very likely others as well.
I should imagine that inertial systems would tend to drift as time went on due to their continuous nature. It makes sense to use GPS to zero out the error every so often.
I wonder if they use libpng too?
>I wonder if they use libpng too?
It seems…unlikely. Who are they gonna display the images to? Passing squid?
How are the squid supposed to know what to proxy forward to the mothership if you don’t display it to them?
You know, it didn’t even occur to me that life saving standards could possibly be proprietary especially when tax money was spent on making them until your previous post on the topic. I’m not naive to that kind of stupidity, but I failed to adequately conceive its pervasiveness. I’m really glad that the frustration paid off in more ways than improving GPSD. These kinds of alarms need to be heard loud and clear and no, it’s not just people getting sore over having an inordinately hard time doing geeky things with GPS units.
“Making standards for life-critical systems proprietary is not just stupid, it’s evil.”
Make that “Standards in general”.
If you really need to get paid to develop a standard, then you should get paid in full before the standard comes into force. The current practise is simply to levy a tax as the population is forced to use the standard and therefore forced to pay.
>>Make that “Standards in general”.
Meh.
If they’re not life savers, it’s just plain stupid and counterproductive… but it’s evil if someone’s life depends on it…
@iajrz
“but it’s evil if someone’s life depends on it…”
Where do you draw the line? Something not working for a consumer generally means it won’t work for a fireman or MD neither.
Anyhow, those setting the standards are quite willing to force those not round the table pay the costs. I consider that evil. But your thresshold might differ.
Meh. You probably could. The information I’ve seen isn’t classified, and as long as I don’t reveal too many implementation details, I’m not violating the ITAR restrictions, either.
I’ve been surprised at how much information about existing military hardware is actually public knowledge.
Making standards private is stupid and counterproductive. Making people pay for those stupid and counterproductive private standards with taxes extracted by force is evil.
If you choose to let your life depend on a stupid, counterproductive, private standard, that’s not evil. It’s just stupid^2counterproductive.
…And sometimes there isn’t any choice. The stupid, counterproductive, private standard is all that is available because it was made a standard by the evil of taxes extracted by force.
Which makes it very clear where the root of the problem exists.
> It seems…unlikely. Who are they gonna display the images to? Passing squid?
Well, squid are quite intelligent and a bunch of trained squid might come in handy in a number of situations, showing them (animated) images would probably the best way to give them commands.
I am interested in AIS (I have an Android app that picks up from one of the open networks) and very interested in disaster-management porn, in fact; thanks for the link.