Unbelievably frustrating that they either don't send, or the call centre doesn't send, the estimate of the signal error. I've tried my best to talk to the various manufacturers about this, and got no results.

It's as if they designed it without consulting SAR at all.

PLB doesn;t work quite as well as that.
A lot of PLBs don't have a GPS. Those that do have the following issues.
Because it's off most of the time it needs to download the almanac (no assisted GPS) and it can take 20 minutes to derive the first GPS signal.
Even when it gets a lock, it truncates the accuracy of the fix in order to transmit the information over the low bandwidth link to the satellite.

This was the experience we had a few years back
http://blog.oplopanax.ca/2011/08/per...s-perspective/

HOWEVER, all that being said I still recommend a PLB over a SEND for most people - since you only use it in an emergency and it is a registered, internationally recognized standard for emergency rescue.

Wow! The variation in PLB points is not good at all.

Oplopanax, your blog is great.....very informative.

That is a very cool blog, really nicely documented. What was the brand of PLB involved in that situation. For sure a PLB with an internal GPS is essential. I've never seen a higher end PLB or EPIRB (mcmurdo fastfind) take longer than few minutes to download the ephemeris/almanac data but that's not saying it couldn't take longer. I recall, in my marine radio training, the US coast guard evaluations backed that up. I'm trying to find some documentation on that.

In EPIRB/PLB testing i've been involved with (marine mostly) the truncation errors never amounted to more than 20m of resolution, but in that particular situation in your blog perhaps there was extensive obfuscation of the GPS satellites? Did they actually turn off the PLB?

At any rate, it makes total sense to always carry some small pen flares (we do).

Again, very cool blog!

Can't remember the model of PLB, possibly McMurdo

They were in very heavy rain, in a mostly east - west valley so GPS signal was probably as bad as it was going to get. This also affects the speed at which the ephemeris will download, as well as the time to get a fix. The other consideratons are where they placed the PLB, of that I am not sure.

I've been using the wrong word to describe the cold start, time to first fix process by describing it as "downloading the ephemeris". When a GPS has been off for a long time it needs to download the almanac which is transmitted every 12.5 minutes, and can takes 12.5 minutes to download. I'm guessing the worst case to download this is about 15 minutes if you miss the beginning of a transmission and your algorithm can't mash two almanacs together.

If the signal is weak or there is an error in the download it could take longer of course.
https://en.wikipedia.org/wiki/Time_to_first_fix

You're right about the truncation, it would amount to approximately 20m - in this case that would still have given us uncertainty as to which bank of a confluence of three creeks they were on!

I'd think all the signalling devices manufacturers should take heed of the kind of situation you describe in your blog, which is an excellent precis of a real event.

It would be interesting to hear from the Dept of Defence, who register and oversee these matters. There are perhaps metrics kept at CFB trenton on these things, though they're probably classified :-(

I think the main issue, which is addressed via education and discussions like these, is that the companies over sell the capabilities of all devices.

From the declarations of GPS accuracy (which I dealt with here http://blog.oplopanax.ca/2012/11/your-gps-is-lying/ ) to either over stating, under specifying or otherwise not managing people's expectations on how a device that you need to reply on works.

Somewhere in the mix of advertising and profit seeking there's a break in the responsibility these companies have in properly informing people about the technology they are selling.

good point. I've got some contacts in coast guard/defence that are into these things in a very official way. I'm going to press them a bit and see if there's a way to get access to their findings