Hacker News Reader

Comment by toufka | Hacker News Reader

What happens when a GPS satellite fails? Or if you have an earthquake? Seems like precisely in those times of national need that one's most sophisticated technologies might be sorely out of commission. A seemingly pretty nice way to inflict 'collateral damage' on one's self. Especially if such a precautionary set of rules were put into more and more machines.

View on HN

Consumer GPS devices have some interesting lock-down rules built into them. They shut themselves down if they detect you are travelling around Mach 2 or above 18000 metres altitude.

The reasoning is fairly obvious - they dont want to see dumb missiles getting upgraded to a DIY guided ballistic missile thanks to a $100 Garmin gaffa taped to the side and a cheap flight control computer.

http://en.wikipedia.org/wiki/CoCom

They shut themselves down if they detect you are travelling around Mach 2 or above 18000 metres altitude

The intention was for a GPS device to disable itself if it was traveling faster than 1k knots AND above 60k feet, not traveling faster than 1k knots OR above 60k feet. Unfortunately, many GPS devices improperly implement this as an OR condition, rather than an AND condition. This becomes problematic when using a GPS in a high-altitude weather balloon to determine altitude and location, which a few friends and I did a couple years ago. We had to be very careful when selecting the GPS we used to avoid this pitfall.

Interesting, I did not know that. Which GPS did you end up using?

Which GPS units are acceptable for this? I'm working on a HAB and I'm worried about this CoCom bullshit. The one I'm thinking of uses a SiRFstarIII or IV; any idea of those use || instead of the proper &&?

Wikipedia article on CoCom links to a post that links to this page:

http://ukhas.org.uk/guides:gps_modules

(Apologies for the untimely reply)

We used a UBlox5, which we purchased from diydrones.com. This was in 2009, so that particular model is now outdated. Both Adafruit and SparkFun now sell devices which are explicitly listed as being HAB-friendly. There's also good guide at http://showcase.netins.net/web/wallio/GPSrcvrsvs60kft.htm. That page states that SiRF-I and II are known to fail above 60k feet, and that some SiRF-III-based receives fail, while others don't.

(If you happen to be in the Bay Area, I'd be happy to meet up and give you guys some pointers sometime; drop me a private email.)

How about the GPS inside Android/iPhone?

Now I have this image of someone throwing a million dollars multi-ton machining center at mach 2 on a ballistic trajectory at his neighbor in the perfect combination of act of war, artistic performance, rich kid play and geeky curiosity ("do the GPS in this thing has the lock?").

> What happens when a GPS satellite fails?

The receivers use other satellites. They have to anyway: You can't get a fix with only one satellite.

I would imagine you normally need three. If I'm visualizing the geometry correctly, the intersection of three spheres is two points, and it should be easy to guess which of the two is the more likely location of your device.

Because relatively cheap GPS receiver clocks drift, typically 4 satellites are needed for a position. You can assume an elevation, especially at sea level, and solve a position with three satellites but this introduces other factors (ellipsoid vs geoid, satellite constellation geometry, etc) that can make a good position difficult to obtain.

In practice there are enough satellites that these considerations are theoretical. Most situations that would involve limited satellites also introduce stuff like multi path reflections that further degrade position accuracy.

Here is a graphic to add to that explanation if anyone is curious: http://gis.stackexchange.com/questions/12866/why-does-gps-po...

In North America your GPS probably connects to between five and twelve satellites.

The interlocks aren't permanent - they can be bypassed by the manufacturer.

I'm not sure you actually meant it as a joke, but that's hilarious in a cynically-detached kind of way. So these machines are not bought but licensed and they will fail as soon as the manufacturer becomes unable to continually grant usage rights, i.e. during a region-wide catastrophe.

I'm merely pointing out that GPS outage would be no more than an annoyance. .

If the machinery experiences significant acceleration in an earthquake, it's not necessarily intact or safe to operate afterwards. The enclosing machine shop is also not likely to be in any state to continue operations if something as large as the pictured CNC machine has been thrown across the room. I don't buy that this issue is a big deal either.

From a software freedom standpoint, yes, this is terrible. But I don't think it's as much of a problem for the owner as the parent suggests.

I have no idea how many of these machines are in a given area, but assuming it is an IMU [1] triggered lockdown, an earthquake could lock down many (10?, 100?, 1,000?) in a single geographic area very, very quickly. That seems like a big problem.

If the lockdown mechanism is GPS, then it should be clear where the actual machine is, so there shouldn't be any need to lock it down unless it's in a black zone.

And what happens if the manufacturer goes belly-up?

[1] https://en.wikipedia.org/wiki/Inertial_measurement_unit

Reply on news.ycombinator.com