aerostitch / testnavit

0 stars 0 forks source link

New Idea to get better travel time calculation driving through cities #201

Open aerostitch opened 11 years ago

aerostitch commented 11 years ago

Issue migrated from trac ticket # 1162

component: core | priority: minor | keywords: eta traffic speed

2013-08-19 13:06:43: geer@pfaffenhofen.de created the issue


Hello everyone,

I was realy impressed, how accurate navit calculates travel times! But when you pass through a city, the accuracy is gone.

So I thought a lot about the difference between a road crossing a small village and the same road crossing a city. Both roads have a standard speed of 50!!

But in the city, the most time is lost standing in front of a red traffic light.

So my idea:

Is ist possible to count the traffic lights standing along the calculated route and then add a variable Delay for each traffic light, to get a better result for the needed time to cross the city?

What you are thinking about such an solution?

geer

aerostitch commented 11 years ago

2013-09-01 03:25:26: usul

aerostitch commented 11 years ago

2013-09-01 03:25:26: usul commented


Hi, thanks for submitting this issue :)

From what I can say, I think it's possible, of course. We have the traffic light elements and I guess we can somehow extend the routing graph, to take the number of lights/crossings that had to be passed into account.

From my perspective, this is slightly related to #1047 (and maybe a dupe?) and a first step towards #570. Anybody else from the team opinions on that?

As we are currently try to find our way back to a stable version, I will park it for only for the next major release.

aerostitch commented 6 years ago

2018-01-21 13:21:27: @mvglasow set severity to normal

aerostitch commented 6 years ago

2018-01-21 13:21:27: @mvglasow commented


1047 looks very similar.

However, #570 I would consider to be a different issue: Here we are talking about traffic lights, which are part of the static map. If we simply add a penalty for each traffic light passed, that penalty (and thus the travel time, as well as the choice of route) would never change, unless the map itself is changed (e.g. the user downloads a later version in which traffic lights have changed). In contrast to that, #570 is about considering the current traffic situations (mostly traffic jams—though delays waitign at traffic lights could also be represented). Traffic changes over time, to the point that calculating the same route twice at different times could yield different results as the traffic situation may change.

Technically, the challenge here is to add a cost to a point (currently only segments have a cost in Navit). For #570, the challenge is to obtain traffic data, translate them into map segments and keep that information up to date.