Route planning estimated time problem?

Started by neatwheat, May 26, 2020, 12:42:04

0 Members and 1 Guest are viewing this topic.

neatwheat

I've got some strange problems with the route planner:

1. the estimated time shown on bottom bar shows somehow wrong values. For example: for a 6km mountain-hike with 1700m elevation it estimates 16 min, when I go to "travel time" it estimates 4:50h. (Planning profile is set to hiking).

2. On the other hand when I plan the same route from the peak downwards, it's the same, but now "travel time" also doesnt work correctly it estimates 1 min for the whole trip down.

Am i doin something wrong?
  •  

poutnikl

#1
Possibly related to this https://forum.locusmap.eu/index.php?topic=6921.msg59879.msg#59879

When I generated the route Cortina to Tofana Di Rozes And Back directly by BRouter application, I got in the result stats about 6:40h and 3:50 or like that. AFAIK BRouter returns ETA to Locus.

Try the same with your affected route, if the issue is Locus ETA or BRouter ETA.

Sent from my Xiaomi MI A2 / Android 10, via Tapatalk

neatwheat

#2
hmm yeah that might be related, seems like something is off here.

I tried the route inside Locus with BRouter and GraphHopper, and there seems to be no change to the values. Route-planner bottom panel time is always massively wrong and travel time has at least a realistic value for ascent - but same route as descent always shows only 1min for whole trip down.

Edit: I just tested it on PC in my Emulator, and the wrong values on the Route planner bottom panel seem to be okay there, so a setting on my phone seems to break it. The broken time for the descent shown in travel time still exists there.

Edit 2: also noticed that after setting a shaping point in certain areas, the summation of the additional time seems to become incorrect on RP bottom panel time
  •  

neatwheat

#3
*Update

1. The completely wrong estimated time in the routeplanner bottom panel (e.g. 16min for a 1700m altitude ascent) seems to be fixed after a preference reset in Locus.
It still shows weird values in some situations or sometimes after adding a second point, but I guess it's like that for everybody atm.

@poutnikl
It seems to me that the est. time in rp bottom panel comes from the chosen routing service (Brouter, Graphhopper) and the values in "travel time" from Locus? Because travel time seems to always stay the same.
  •  

poutnikl

#4
Quote from: neatwheat on May 28, 2020, 11:07:25

@poutnikl
It seems to me that the est. time in rp bottom panel comes from the chosen routing service (Brouter, Graphhopper) and the values in "travel time" from Locus? Because travel time seems to always stay the same.

Yes, that is correct, also confirmed by @menion . Bottom bar value may be from Locus as well as some default internally calculated value, if route provider does not return ETA info.

AFAIK, BRouter use some internet formula for slope aware hiking speed, and a constant power bicycle kinematic model for bicycle ETA ( in recent built-in profiles, not yet in my profiles ).

I have realized 85W constant power provide ( for my total mass and travelling habits/performance) quite good  long time average effective travel speed.

Sent from my Xiaomi MI A2 / Android 10, via Tapatalk

neatwheat

#5
Thanks for clarification! Then it definitely seems that Locus sometimes has problems taking the correct values from Brouter - on the same route Locus sometimes adds time when adding a second waypoint and sometimes it doesnt.

Btw I'm also trying out your SAC6 profile from Github - is that still recommended or obsolete?

Edit: Sorry didnt see your Edits before answering ^^
  •  

poutnikl



Quote from: neatwheat on May 28, 2020, 11:33:28

Btw I'm also trying out your SAC6 profile from Github - is that still recommended or obsolete?

It is getting aged and definitely may deserve some updating ( rather the hiking profile template in the other repository ).
But it is still valid, not obsolete.
Note that I did not have opportunity to check it in last year's in terrain.



Sent from my Xiaomi MI A2 / Android 10, via Tapatalk


neatwheat

Quote from: poutnikl on May 28, 2020, 12:20:54
It is getting aged and definitely may deserve some updating ( rather the hiking profile template in the other repository ).

Thanks for the hint, I'll try the template too!

Quote from: poutnikl on May 28, 2020, 11:11:30
AFAIK, BRouter use some internet formula for slope aware hiking speed, and a constant power bicycle kinematic model for bicycle ETA ( in recent built-in profiles, not yet in my profiles ).

I have realized 85W constant power provide ( for my total mass and travelling habits/performance) quite good  long time average effective travel speed.

I'll keep that in mind for the next bike trip!
In terms of hiking the BRouter Internet formula unfortunately seems to not be very sophisticated. I often tried calculators based on the Naismith Rule for trips in the past and it seemed to produce a better "rough" ETA.
  •  

poutnikl

Quote from: neatwheat on May 28, 2020, 15:34:40

Thanks for the hint, I'll try the template too!

If experimenting with the template and if a little familiar with Linux, you may find the use for the sedbatch bash script in Brouter-profiles, that can be run in Android termux Linux console, generating profiles directly from the local or GitHub template. You may want to taylor it for your own modifications.

There is also the original Windows batch sedbatch.bat, but rather obsolete and much less sophisticated.



Sent from my Xiaomi MI A2 / Android 10, via Tapatalk

neatwheat

Quote from: poutnikl on May 28, 2020, 20:29:56
If experimenting with the template and if a little familiar with Linux, you may find the use for the sedbatch bash script in Brouter-profiles, that can be run in Android termux Linux console, generating profiles directly from the local or GitHub template. You may want to taylor it for your own modifications.
Thanks a lot, but I fear the script might be too much for my small skillset ...

*Update
Toogling "Include navigation hints" on or off also changes the ETA from RP bottom panel massively
  •