Driving duration comparison with Google and OSRM

Directions API Route Optimization API
1

I was doing a comparison of distance/duration between Google, Graphhopper and OSRM. Below are some of my observations and questions.

  1. The distance values are quite similar between Google, Graphhopper and osrm with negligible difference. But with duration I observed that Graphhopper underestimates driving duration with a difference varying between 1-6 minutes compared to google. I did this comparison for about 500 different locations in Australia. This time difference was observed for distances in the range of 10-40km. For a short driving distance of 10kms, a time difference of 3 minutes will make a huge difference. I also came across this post from 2016 : Driving duration compared to Google

  2. I was using the trial version of Graphhopper. Are the driving duration/distance values returned same as with the paid account?

  3. For the driving duration in Graphhopper, do we get the average duration or the duration in peak traffic conditions. Can we specify the time of the day? In one of the blogs I saw that we can specify the traffic information by changing the data provider to tomtom. Is there any other way ?

  4. Also, the Graphhopper driving duration was always more than the osrm duration.

Kindly share more details on the underestimated driving duration when using Graphhopper.

2

No, they are not.

It does only make sense if you use our data provider TomTom. There is no other way to consider historical traffic information.