Ascent calculation is inaccurate

Posts for this topic...

• Post your comments.... Sign In to Post
• Andrea Ticci Monday 28 Mar 2016 18:34:44

  Thank you John,

  You have done a fantastic job with this site and I understand there is a limit to what can be realistically done calculating elevation from map.

  Just to clarify what I was meaning,I am providing what seems to be a good example:

  This route I have generated with plotaroute https://www.plotaroute.com/route/187663 is quite famous in the area. The route is a continuos climb with not a single meter going downhill over the entire length (13km). The starting point is at 154m above sea level, the end is at 814m above sea level.

  The real elevation gain for this route is 814 - 154 = 660 m (source: google earth, in good agreement with plotaroute values and altimeter readings)

  Plotaroute elevation profile for this route https://www.plotaroute.com/routeprofile/187663 shows a fragmented pattern with continuous up and down that do not exist.

  The unfiltered elevation gain from plotaroute provides 1305 m  that is about twice the real elevation gain.

  I believe the limit is provided by the maps used to generate the elevation profile.

  You are right in saying that barometric GPS absolute elevation readings can be inaccurate especially if you do not calibrate the altitude with a known value at the beginning of the ride to eliminate drifts due to barometric pressure variations (sometimes I read negative altitudes at sea level). In any case the differential elevation gain remains quite accurate (e.g.  My GPS with barometric altimeter is quite in good agreement with the  calculated 660 m elevation gain of the above route, the deviation from  the 660 m is always less than 4% ).

  I would suggest when uploading rides to plotaroute, to provide the option of recalculating the elevation data or to leave the data in the uploaded route.

  Regards

  Andrea

   

• Nisse Son   Monday 28 Mar 2016 19:20:09

  John, I first of all must give credit to how you and the Plotaroute team takes on customer complaints, issues and feedback. You dive into things fast and with engagement, this is an area where you stand out positively from a lot of competitors!

   

  Hi Olof - I've had a quick look at your Sa Calobra route, which has a total ascent of 914m on our calculations but 715m if you assume there are no downhills.  However, it does appear that there are some downhill stretches, at least according to our data but also according to Google's data.

   

  Trust me, there are none. I would not use hypothetical examples when I post my findings, I have ridden all of them. If in doubt, post a question to any of the bigger road cycling forums and ask how much downhill it is on the first 9.9km of Sa Calobra. :-) The total from Plotaroute is ~200 meter of.

   

  I'm not saying that our data is perfect, as no solution is perfect and there is a limit to what we can provide as part of a free service, but it does seem like our data is fairly consistent with Google's.

   

  This is where the problem arises, you assume Google's data to be accurate. It is not when it comes to small roads climbing up and down mountains. The simplified model they use doesn't give correct data for this use case when used without smoothing of it. Lets take the part between 6-7km of the Sa Calobra as an example: https://www.plotaroute.com/route/187724 What according to Google is an undulating road is not, this is an error due to their simplified model of terrain. Again, I have ridden this road and know where there are and are not downhill. Do download the .kml and look at the 1 kilometer stretch in Google Earth street view, and you will see for your self that there are definitely no downhills to speak of. (The slope is quite the same all way through, which makes sense as the road is built to accommodate for the tourist buses and therefore is about 6-7%). 

   

  This is not a problem with these specific roads, but a general issue. For 95% of roads the error is so small you won't notice, but please remember that it is for those last 5% where a service like this is used a lot.

   

  Pleas bear in mind that this is a free service and there is a limit to what we can realistically do.

   

  Plotaroute is not a free service, you work in Freemium-model. (I'm happily paying for the service.) I for one think it is an important one for how I use the service. But I fully agree on that you need to make decisions on what bugs to fix vs. what features to build and this might not be the most important one.

   

  What I and others propose is not that you try to find correct data for all roads in the world, but look into how smoothing of the data would give more accurate information for total ascent and for the time estimations. 
• plotaroute admin   Tuesday 29 Mar 2016 10:37:27

  I'm afraid we don't have the resources to keep doing detailed analysis of individual routes.  If Google's data isn't giving you the accuracy that you are looking for we have little chance of bettering that!  plotaroute is a tiny operation and is offered as a free service.  Yes, we do have an ad-free option but the nominal annual fee for this just offsets lost advertising income - our total revenues are not enough to pay the salary of one person let alone a whole team!  The project is very much a labour of love at present that I hope will grow into something bigger, so please bear with us, as we can only achieve what is feasible and within our means.

  I understand the desire to try to get our Total Ascent figures to match those from other systems, but my main concern is to ensure that people can compare routes that we have on plotaroute on a like for like basis, so I would not want to apply different rules for different routes, for example by sometimes using GPS elevation data in place of the data we would normally use.  Also, we would need to make fundamental changes to the system to do this, as elevation data is not captured and stored on a route by route basis at present.

  I think one way forward would be to calculate two figures for the Total Ascent - a 'raw' figue (the current one) and a 'smoothed' figure (sampling elevation readings less frequently).  This will retain the integrity and comparability of the existing routes while provding an alternative figure that reflects the degree of smoothing that is probably found in other applications. I can add this to our Feature Request list if you feel this would be helpful, where anyone interested in it can vote for it.  As you'll appreciate though, there are many other users of plotaroute who are asking for enhancements and new features, so I do need to take into account which ideas are most popular.

  John

   

• Nisse Son   Wednesday 04 May 2016 20:58:57

  "I'm afraid we don't have the resources to keep doing detailed analysis of individual routes.  If Google's data isn't giving you the accuracy that you are looking for we have little chance of bettering that!"

  I don't think anyone expects you to, but you could use what the other services do, heck you have already built it yourselves! :-) The Google data doesn't have the accuracy to be used for this application as raw data, it has to be smoothed a bit. This is what other services do to make the accumulated numbers more true to the real world. Without this, the totals can be off by 50%, as previously shown.

  But good news, there is an easy way to solve this! What I (and possibly others) ask for is just that you use smoothing for the data. Plotaroute have the feature already in the "Route Profile Tool". If I use this for any of my done rides and set "Filter Bumps" to 10.0% and "Elevation Intervall" to ~200 meters, both the hills inclination and total data matches reality. If e.g. these numbers would be default things would be as I (and possibly others) would like it to be. 

  Not only are the totals off as of today (annoying, but still just numbers), but the Route Timer-estimation gets so wrong it is useless. If the Route Timer and Download features used the smoothing from the "Route Profile Tool" instead of raw data, then those feature would be much more useful.

  So, my suggestion is: make some smoothing default (if not for all users all the time, but could we please have it on a user base?). This will solve this problem to at least the use cases I have seen in this thread.  

• plotaroute admin   Friday 06 May 2016 09:50:14

  I'm happy to look at smoothing the elevation data a bit to reduce the total ascent figures  if enough people feel this would be helpful.  I've added it to the Feature Requests list (number 52), so that anyone interested in this can vote for it.  At the moment only 3 people have asked for this on the posts below and we have over 50,000 registered users on the site, so I need to consider that some people may prefer that the data wasn't smoothed.  My personal opinion is that it would probably help for the Total Ascent figures to be more comparable with other applications and, so a bit of smoothing is probably a good idea, however  I think I would only want to increase the sampling interval from 30m to 90m.  This would also mean having to recalculate the Total Ascent figures for every route on the database, so it would be quite a big task.  And there would be additional development work if we also smoothed the actual elevation profiles, but if smoothed data is used to calculate the Total Ascent figutes it would probably be best to use smoothed data for the elevation profiles too. 

  John

   

• Claudio Crivelli Wednesday 11 May 2016 09:40:17

  Reagrding ascent calculation I have to say estimation by Plotaroute is quite good during the editing. But once I saved the track, I see even a +40% increment without any reason.

  If I go back to editing the same track and I just move the start-stop point by few meter, the system updates the total ascent and the number comes back to "real" values. So to me the problem seems to be related to the SAVE process... Even without saving, if I switch to the profile tool I see the increased value... It seems the right value is given only during the editing and only in the "HILLS" area (below the map).

  Is this happening only to me?

  Thanks,

  Claudio

  ps: plotaroute is a great tool! many compliments!

• plotaroute admin   Wednesday 11 May 2016 12:48:16

  Glad you like the site Claudio - thanks.  

  The Total Ascent figure shown on the elevation profile as you are plotting a route is just an estimate based on a small sample of elevation readings (a maximum of 500 at equidistant intervals across the route)  - a more detailed calculation is made when you save the route using elevation readings at 30m intervals, as it isn't feasible to recalculate this in such detail as you are plotting.  We do adjust the 'as-you-plot' figure by a factor to try to compensate for the smoothing effect of using elevation readings at bigger intervals, but the impact of using bigger intervals differs according to the amount of bumps and dips on the route, so it will sometimes vary quite a bit from the final figure.

  John

• Pierre Chartier   Saturday 21 May 2016 17:44:45

  Hello

  I think this would be a nice idea, not to make it comparable with other applications but also to make it more accurate.

  I leave in a mountain area (French Alps), so I ride frequently rides with a lot of elevation. The total elevation on plotaroute is often at least 50% too high. Example : I rode yesterday https://www.plotaroute.com/route/216481

  Total ascent is 2671m on plotaroute. Yesterday a friend and I measured a total ascent of around 1820m (we had two different GPS)

  I found similar results on pure uphill rides when comparing also with the altitudes written on a map.

  The percentages as well are too high. I frequently get on plotaroute a max percentage of more than  35% (38.8% on the ride above) which is impossible to climb on a bike (at least for me!)

  Keep on the excellent work you are doing. Your site is great!

   

• plotaroute admin   Monday 23 May 2016 11:12:28

  Thanks Pierre for your feedback - glad you're enjoying using the site.  Adding some smoothing to our elevation data is on the Feature Requests list but please vote for it to move it up the list (number 52).  Please be aware though that GPS devices are not a reliable bechmark for ascent calculations - we've looked at numerous examples of raw GPS recordings and usually find big gaps in the data (probably due to poor signal reception) that result in ascent figures being understated.

  John

• Pierre Chartier   Monday 23 May 2016 18:10:06

  Thanks. I voted or it.

  There is just near my place an ascent which is pure uphill on 18km. It's about 1400m ascent. Next time I climb it I will see what plot a route gives and compare it both with a map and with what my GPS gives.

  I will let you know.