Forerunner 955 GPS Problem

The first track must have completely lost GPS lock. On the way back it must have picked up just enough signal to plot those seven inaccurate points.

GPS accuracy drops when the angular spread of satellites that can be received is low.

Hi! I saw your results. Great half marathon.))
Noticed that the gps broke at the very end. Unfortunately, I have no change either.

For some reason one run after the firmware update recorded fine. After that, all runs were in the old scenario. I think you saw them on my Garmin Connect.

Nevertheless, walking and biking recorded perfectly. And that's really weird.

It turns out that this problem is only present for jogging/running. To be honest, I'm tired of hoping for something. I'm going to change my watch for another one. This is very sad.

By the way, if they say parking lots, dense woods, cloudy weather - it's complete nonsense. It can be a problem, but as you rightly pointed out - one way it's ok, the other way it's broken there. I've been experiencing this kind of thing for almost a year now. There is no answer to this problem. We can only hope that the next firmware fixes it. But I would recommend not to wait, but to go and change the watch. Even for the same one. Maybe it's a defect of exact your and my watches.

Thank you :) 

My personal opinion is that this is a software problem. In other words, this is a bug that the developer team has made in the algorithm used to combine the received GPS signals. They may have even said that if the GPS signal received every second is correct on one side and wrong on the other, if there is an access problem in the GPS signal received every second, take the nearest GPS point as a reference and create a plane accordingly. They can fix it in the next release. I don't even want to think that Garmin Developers can't solve a simple software problem, so I will continue with the 955, the values that my watch has given me so far are more than what I get. :)

Normally GPS data is sanitized using something like en.wikipedia.org/.../Kalman_filter

You are absolutely right. This is a code error. For some reason we are the ones with this error. I don't know. I have no idea. This error will be fixed in the next firmware. It's possible. Or maybe not. The developers have been promising to fix it since last September. But so far there are no results. I hope you and I will have enough patience to wait for it. But my patience is running out.

In fact, my personal opinion is that this is probably the right way to have the GPS track, because when the GPS signal is lost, there are 2 alternatives; either to combine the gap between the last GPS track that follows the path and is recorded after the signal is lost and the first GPS track that is received when the signal is recaptured, which may not always be accurate, or to take a reference and draw the most appropriate route according to this reference. I am also a developer. So these exceptions are simple and fixable exceptions. For the Garmin developer team, with all the improvements they have made with the hub sensor, I think GPS problems like this are child's play.