# Too long travel times - calculation inaccurate (too high) **URL:** \ **Category:** openrouteservice\ **Created:** [August 1, 2024, 9:21am UTC](https://ask.openrouteservice.org/t/too-long-travel-times-calculation-inaccurate-too-high/6368 "2024-08-01T09:21:11Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![zacki41](https://ask.openrouteservice.org/letter_avatar_proxy/v4/letter/z/4491bb/32.png) [@zacki41](https://ask.openrouteservice.org/u/zacki41)\ **Post date:** [August 1, 2024, 9:21am UTC](https://ask.openrouteservice.org/t/too-long-travel-times-calculation-inaccurate-too-high/6368/1 "2024-08-01T09:21:11Z")
Dear ORS Team, I am using ORS QGIS Plugin and am running my own ORS V8 instance. When I calculate traveltimes I experience huge differences (too long travel time but correct km distance) between ORS calculations and other engines (Valhalla, Maps, Openstreetmaps). I have checked it with you demo instance, where I also receive the same results. Let me give an example: > **[OpenRouteService route planner - directions, isochrones and places](https://maps.openrouteservice.org/#/directions/Bezirkskrankenhaus%20Lienz,Lienz,TR,%C3%96sterreich/Rotes%20Kreuz%20Silian,Sillian,TR,%C3%96sterreich/data/55,130,32,198,15,97,4,224,38,9,96,59,2,24,5,192,166,6,113,0,184,64,70,0,152,3,160,29,128,54,0,89,12,188,128,105,110,32,14,1,152,5,96,1,128,78,82,6,226,56,181,22,53,40,55,38,82,169,124,44,65,209,1,0,3,170,120,17,19,97,202,0,23,148,0,182,185,243,209,0,186,4,0,102,240,0,219,165,194,22,52,120,0,220,144,7,48,11,70,25,52,89,71,163,165,62,151,209,12,26,207,23,210,7,71,93,17,22,29,22,4,0,23,222,40,0)** > > Openrouteservice is a open source route planner with plenty of features for car, heavy vehicles, hiking, cycling and wheelchair returns 46 minutes Openstreetmaps: > **[OpenStreetMap](https://www.openstreetmap.org/directions?engine=fossgis_osrm_car&route=46.835%2C12.765%3B46.747%2C12.423#map=12/46.7914/12.5938)** > > OpenStreetMap is a map of the world, created by people like you and free to use under an open license. Returns 32 minutes and 32 km and Maps "[Google Maps](https://www.google.at/maps/dir/A.%C3%B6.+BKH+Lienz,+Emanuel+von+Hibler-Stra%C3%9Fe,+Lienz/%C3%96sterreichisches+Rotes+Kreuz+Ortsstelle+Sillian,+Sillian+185h,+9920+Sillian/@46.7908938,12.5118156,12z/data=!4m14!4m13!1m5!1m1!1s0x47775d8a7be69a97:0xf80b1cac4efb82a4!2m2!1d12.7642246!2d46.8354816!1m5!1m1!1s0x4779d6717bd1be15:0xd0101d79f499e316!2m2!1d12.4232552!2d46.7471497!3e0?entry=ttu) (Bezirkskrankenhaus Lienz to Rotes Kreuz Silian) 30 Minutes This happens with a variety of cases, all displayed here [BVRD Transparenzlandkarte RD](https://transparenzlandkarte.bvrd.at/?k=90a34d334) - (The number expresses the minutes to shortest next station from a matrix calculation) Almost all are too high. Would be very grateful to get some insights what could be the problem. Florian ---
**Author:** ![jschnell](https://ask.openrouteservice.org/user_avatar/ask.openrouteservice.org/jschnell/32/971_2.png) [@jschnell](https://ask.openrouteservice.org/u/jschnell)\ **Post date:** [August 13, 2024, 1:45pm UTC](https://ask.openrouteservice.org/t/too-long-travel-times-calculation-inaccurate-too-high/6368/2 "2024-08-13T13:45:35Z")
Hey, this is probably due to the acceleration heuristic, used to improve driving times in urban areas, that seems to be the culprit here. Have a look at this [issue on driving speeds](https://github.com/GIScience/openrouteservice/issues/622) that has a bit more information. Best regards ---
**Author:** ![zacki41](https://ask.openrouteservice.org/letter_avatar_proxy/v4/letter/z/4491bb/32.png) [@zacki41](https://ask.openrouteservice.org/u/zacki41)\ **Post date:** [August 15, 2024, 7:32pm UTC](https://ask.openrouteservice.org/t/too-long-travel-times-calculation-inaccurate-too-high/6368/3 "2024-08-15T19:32:22Z")
hi @jschnell - can you hint me towards a solution - I’m using the docker [Running with Docker | openrouteservice backend documentation](https://giscience.github.io/openrouteservice/run-instance/running-with-docker) I fear that the `ACCELERATION_SPEED_CUTOFF_MAX` can only be changed when building it from source and changeing the cutoff max correct ? [Building from Source | openrouteservice backend documentation](https://giscience.github.io/openrouteservice/run-instance/building-from-source) ---
**Author:** ![jschnell](https://ask.openrouteservice.org/user_avatar/ask.openrouteservice.org/jschnell/32/971_2.png) [@jschnell](https://ask.openrouteservice.org/u/jschnell)\ **Post date:** [August 16, 2024, 7:18am UTC](https://ask.openrouteservice.org/t/too-long-travel-times-calculation-inaccurate-too-high/6368/4 "2024-08-16T07:18:09Z")
Hey, you may try setting `use_acceleration` to `false`, see [the encoder options documentation in our backend docs](https://giscience.github.io/openrouteservice/run-instance/configuration/ors/engine/profiles#encoder-options). Not sure if that will have the desired effect - if not, you’ll have to look at the source code as mentioned. Best regards ---
**Author:** ![Marcos\_Andrew](https://ask.openrouteservice.org/user_avatar/ask.openrouteservice.org/marcos_andrew/32/3209_2.png) [@Marcos\_Andrew](https://ask.openrouteservice.org/u/Marcos_Andrew)\ **Post date:** [August 16, 2024, 9:35am UTC](https://ask.openrouteservice.org/t/too-long-travel-times-calculation-inaccurate-too-high/6368/5 "2024-08-16T09:35:02Z")
Hi, I think you should facing this issue after long time might be related to the speed profiles or road classifications used by your ORS instance which can significantly affect travel time calculations. ORS relies on specific profiles that might differ from those used by Valhalla Google Maps or OpenStreetMap leading to discrepancies in travel time despite accurate distance measurements. You might want to review the speed settings and profiles in your ORS setup to better align with the other engines or adjust the profiles to better match real world conditions. Thanks ---
**Author:** ![zacki41](https://ask.openrouteservice.org/letter_avatar_proxy/v4/letter/z/4491bb/32.png) [@zacki41](https://ask.openrouteservice.org/u/zacki41)\ **Post date:** [August 21, 2024, 9:51am UTC](https://ask.openrouteservice.org/t/too-long-travel-times-calculation-inaccurate-too-high/6368/6 "2024-08-21T09:51:56Z")
Hi, as of now, it did have the desired effect. Thx. ---
**Author:** ![zacki41](https://ask.openrouteservice.org/letter_avatar_proxy/v4/letter/z/4491bb/32.png) [@zacki41](https://ask.openrouteservice.org/u/zacki41)\ **Post date:** [February 8, 2025, 12:35pm UTC](https://ask.openrouteservice.org/t/too-long-travel-times-calculation-inaccurate-too-high/6368/7 "2025-02-08T12:35:09Z")
Hi again, after upgrading to ORS V9, I experience yet again the issue, despite having set the use\_acceleration to false. My example shows a discrepancy between the to points Point 0: 13.637017, 48.003171 Point 1: 13.716820, 48.004799 ORS V8 with use\_acceleration: false ![grafik](https://ask.openrouteservice.org/uploads/default/original/2X/c/c45eb836ab07e04728045ab95e0289d3661b3e90.png) results in a 8.5 Mins drivetime ORS V9 with use\_acceleration: false ![grafik](https://ask.openrouteservice.org/uploads/default/original/2X/d/d64bd42f7d5221c9d9e1a9b11fe9320ca7d0b83b.png) Results in a 14.28 min drivetime. Tried to validate with Openstreetmaps and it returned > **[OpenStreetMap](https://www.openstreetmap.org/directions?engine=fossgis_osrm_car&route=48.003541%2C13.636994%3B48.004897%2C13.716906)** > > OpenStreetMap is a map of the world, created by people like you and free to use under an open license. 9 mins Drivetime And Maps returned 9 mins as well. In my ORS V9 profile I set the following and rebuilt the graphs. ``` profiles: driving-car: encoder_name: driving-car build: encoder_options: turn_costs: true block_fords: false use_acceleration: false preparation: min_network_size: 200 methods: ch: enabled: true threads: 4 weightings: fastest lm: enabled: false core: enabled: true threads: 4 weightings: fastest,shortest landmarks: 64 lmsets: highways;allow_all ext_storages: WayCategory: HeavyVehicle: Tollways: WaySurfaceType: RoadAccessRestrictions: use_for_warnings: true service: execution: methods: lm: active_landmarks: 6 core: active_landmarks: 6 ``` Is there anything else I can do in V9 to gain results as seen in V8 by disabling accelerate? Thanks Florian