Smoother average for speed value

Wolfgang Kufer shared this idea 4 years ago
Completed

Hi,


I was out for long trip by foot (with a large group of people) and I had to set the speed of the group. I tried to use the 'actual speed' in my dashboard, but the displayed value is much to instable (when hiking). It shows values from 4.4 km/h up to 5.8km/h when walking at constant speed.


Would it be possible, to add a (speed dependant?) averaging filter?

Comments (6)

photo
1

I remember some Java ME GPS software I used in past ( Trekbuddy or GPSMid ) had on their dashboard last minute average. I think it would be great, as aditition to total and while-moving average.

photo
1

Hello Wolfgang and Libor,

interesting observation. On many places, Locus use flltered speed from last 60 seconds. Like when Locus decide if use hardware compass or not, when compute times for guidance or time rings etc.

Anyway on all places where is used dashboard or any extension of same system (like panel for track recording, satellite/compass screen, etc.) is used fresh current value.

Because same value may be used over Locus API in various add-ons, it is not simply correct, to change this current value to averaged value. I plan to do some updates in Locus API later next month, so I hope I won't forget and update this as well. If there will be some varying values in next Locus version, feel free to kick me (remind me) ;).

photo
1

Hello Menion,


thank you for considering this request so positive. Offering a second 'averaged' speed as field value for the dashboard would be a great idea. And would not change the API for other addons.


And having the choice of the integration time of this averaged value would be a very nice extra.

photo
1

Hi Menion,

I would like to express my opinion on average speed

1) STATISTICS of the route we have average speed (generally = traffic + stops) and average speed for traffic. I hope that these speeds are calculated for the entire duration interval and the length of the route, so they are real (?!).

2) If while moving, the average speed is taken from the last 60 seconds, I think that this value is VERY SENSING for me, because:

a) I overcome sections of the route at different speeds, which is influenced by a number of factors (eg wind, uphill driving, or prosaic temporary indisposition). And it is these 60 seconds that reflect our real average speed at the moment, which can be significantly different from the total mean.

b) whereas when we take into account the average speed from the entire route already covered, for a given moment of the route we must understand that this value is affected by longer time intervals, and only in the final phase of the route can we rely on this average to obtain a reliable score.

3) If I had to find fault with it - why 60 seconds? Because maybe this value is variable and depends on the speed ... Maybe it is better to leave it as it is, the more that reaching the finish line temporally coincides with the time given by the navigation, and if we want to know the actual average speeds (if they are), let's look on them in the tab STASTISCTICS route.

Waldek

photo
1

Hello Waldemar,

Statistics in track detail are really computed from the whole track.

60 seconds was chosen long time ago as an experimental value. Anyway, from that time, the system changed a little bit and app internally use an exponential filter that allows quick reactions on bigger changes in speed while still keeping useful filtering results.

Also, mainly information for Wolfgang, in the latest app version has introduced a system for filtering of location (in settings > GPS > Location filter). This filtering also affects speed, so feel free to play with it. Even a light filter should give really good results.

So I believe that this idea is implemented over this filter. Thanks

photo
1

I had the same request to smooth the speed graph and improve the accuracy of the "maximum speed" when fast walking. Due to the noisy speed value, the maximum is always too high.

Thank you for the hint (position filtering). It was disabled. I will test with this filtering next time.


And... thank you for this nice application. I use it every day now.