Track timer mode is very unreliable


#21

@VincentSun is looking into the issue and will also be updating the LapTimer Icons with text labels!


#22

I love this system when the laptimer works. Not so much when it doesn’t. I’ll be testing this again in the next few weeks at another event. With that said, thanks for the constant improvements of the product


#23

Just got back from Lime Rock Park to find that the lap timer only worked once out of four sessions. Every time prior going on track I chose the following:


Last time I had my logs checked by support they stated I was using the “point to point method” which I’m now 100% certain I never use. There has to be some sort of bug causing this.

One thing to note I’m removed the suction cup from the mount and replaced it with magnets to prevent wobbling. Would a strong magnet cause this functionality to work sporadically?


#24

Hi @g.assenza, I’m sorry to hear you had issues with the LapTimer at your last track session. First question is, did you “sync tracks” through the mobile app before selecting “tracks nearby”?

Since the GPS module is built inside the mount, the magnets could be interfering with GPS which is how the LapTimer calculates your lap times and position. You can check the stability of the GPS signal by loading the track session into Waylens Studio, then “export raw data”. Within the CSV spreadsheet you should have GPS coordinate info and you can see if signal dropped out at all…

We also have a GPS signal test within the Debug menu of the Horizon so you can test while at the track to make sure you have good signal.

Just wondering if you’ve ever tried to use LapTimer’s “manual mode”?


#25

Have a look at what I posted in the very first message in this thread. Even with a great GPS signal, your lap triggering logic doesn’t work effectively, because it relies on absolute distances from the finish line, and at higher speeds, particularly when GPS reception isn’t perfect, your camera will never notice it crossed the finish line because it never gets a fix close enough.

You can observe this happen live; before the camera starts recording laps, it’ll tell you how far you are from the start/finish line in yards. When it works, you see it reporting 20…10…0… and it starts recording. When it fails to work, it will be more like: 25, 15, 5, (you cross line here), 5, 15, 25. I made up these numbers, but my point is that the camera got a fix 5 yards before the line, then five after, neither of which was close enough to trigger the lap.

Let’s say you have a track map with the endpoints of a line going through start/finish defined by the points A, B in lon/lat space. Your car’s location is C.

The camera should compute this at every GPS fix:

d = (Cx - Ax) * (By - Ay) - (Cy - Ay) * (Bx - Ax)

Where Cx=car longitude, Cy=car latitude, and so forth.

The value of d isn’t important, but the sign of d is. When you cross start/finish, the sign of D will flip. So, if it goes from positive to negative, or vice versa, you’ve crossed start/finish between two fixes, the distance doesn’t matter. GPS is noisy, so this could oscillate a bit near the finish line, so you have to ignore every sign flip after the first within a window of a few seconds. I guarantee you that this way of triggering won’t miss laps.

(I’ve built these kinds of systems for a living, so the bug in your camera is really clear to me).


#26

Thanks for the explanation @marcinr. I know you have discussed with our product manager and engineer @VincentSun. Ultimately if the feature logic needs to be updated it will require some development work and a FW update.


#27

Hello Sir,
Unfortunately we don’t use 2 gps points to denote a start line, instead we only use one gps point to mark a start or finish line. Therefore the first step of your solution would be to collect the 2 points for each start/finsh line of every track in our DB. It is difficult to mark them on ourselves and it would involve a lot of verifications on the data’s validity as well.
Thank you!


#28

Wow, yeah, then this is an annoying problem.

You could still make triggering better by measuring whether you are getting closer to or farther from that point. When you switch from approaching to receding, you have crossed the line somewhere in that transition, so mark it as a lap start/finish, even if you didn’t get close enough to that point.

Perhaps you could crowdsource map start/finish lines. I’d be very happy to update all my local track maps, since I want them to work well.


#29

Can’t google maps give us GPS coordinates of start/finish lines? Just zoom in to the points and put down a marker and get the coordinates. Or is not precise enough?


#30

@KenYip That would require validation.

@VincentSun What you can do is try to determine what falls on a track line, and create 2 GPS points from adding and subtracting a second in a perpendicular direction to the track’s direction.

I’m not sure how the track data is formatted, but it should be possible to create a perpendicular from the track data. If you need a hand you can send me an email personally.


#31

Thanks guys. I know that a perpendicular line should work much better in determining the cross timing. I will spend some time on the optimizations when I am a little bit free.


#32

I just want to add that I used the Horizon cam for the first time at track event and experienced the same problem of the camera failing to detect start/finish line most of the time. I’m close to abandoning and returning this product if you are unable to find an appropriate fix.

Please keep us posted on your commitment and progress to correct this serious bug.


#33

Hello Sir,
I am sorry for the inconveniences. May I know what was the test mode you used? Manual or the auto ? And what is the track name?
Thanks.