Facebookstreetmap.com

OK, following a personal email from Stephan I’ll attend any meeting with a view to moving forward. I hope that if the Thai mappers can come up with a clear set of instructions to the DWG, maybe we make can some progress.

Everything Tom & Peter say comes from the same heart as me … when we map, its as a human, looking logically at the road network. This is what dedicated OSM mappers like us did … see something wrong, or something that was missing with every new set of aerials … we added it. Smooth a curve here, straighten a road there … continual improvement. The FB team just banged it in, and moved on. Eventually the sheer volume of “tidying” FB inputs was just too much … we never stood a chance.

I note we are being self critical over our local definition of road vs. track … Peter is right… if in doubt, downgrade it, and revert later when better evidence comes to light.
My biggest issue with FB is that clearly they do little to correct the existing map when they import … you know it makes sense if, when you draw a road, take a quick look at the adjacent roads and see if with a few more additions, navigation would make sense, and through routes would be created.
I have always criticised FB from this angle, and once you see a nest of FB imports, there is a tendency to move on … its heartbreaking to see it, and a waste of time to start editing it.

And by way of a real example, I was driving down to Kanchanaburi last week, and noticed a lot of U-turns missing on Hwy 340. I know that in the early days, the aerials were poor and a lot got missed, but in Thailand, passing the road you need and doing a U-turn is fundamental part of the driving here. I vowed to correct on my return…

Oh yes, Facebook got there first, and I thought, well lets just log what “sloppy” mapping I notice, and it didn’t take long to fill a page:

Chai Nat :
Way: 527960964 - VLD007 added many residential roads either side of the Trunk Hwy but completely ignored the link roads (U-turns) that are necessary to turn into them.
Way: 528184019 - VLD007 draws a long road as a residential despite fact its clearly unpaved, and only passes one farm, but then fails to draw the bridge at the eastern end. This should be a track, or at the very least, get an unpaved tag.
Way: 527984298 - Ditto the above, but it doesn’t even pass a single structure once leaving the asphalt road.
Way: 527960922 - The residential road is drawn, but then no attempt has been made to clean up the poorly drawn junction it connects to… the one with the fixme=precision tag clearly visible. Its also unpaved, but no tag was added.
Way: 527963823 - Does the road really need this many nodes in the curved area ? It can’t be like this in real life? I would smooth out, deleting surplus nodes.
Way: 527960904 - The way connects to Hwy 340 Eastbound carriageway, but not the Westbound when clearly there is a U-turn gap in the median to do this.
Way: 527922341 - Again, Im guessing FB automated software picked up the residential road, and VDL007 connected it to the Southbound carriageway. There is a U-turn link that connects it to the Northbound but he was either too lazy or simply blind and did not draw in the link. Its errors like this that makes GPS navigation useless.

It’s just the whole “bang it in attitude”; roads are split in random places, ever heard of the merge command, FB ? And crossroads are a rarity to them … take a look at the number of residential roads where the connection is clearly a crossroads in reality, but FB just goes with wherever their automated system puts the node… even if its is only 5 mtrs offset. So whats the issue ? I’ll tell you, my GPS says “turn right, then turn left” and then I’m heading down the wrong road. But then they never ride here, and with a “successful OSM project finished” on their resume, off they head to another posting.

And so with heavy heart, they say “we are really concerned, but VDL whatever has left us now” … OK, if he was that bad, revert every one of his edits. Some of these date to 2017, so clearly going back and putting things right doesn’t seem to be working. Few changes are taking place.

And while I’m on the subject, ask VLD012 why he changed Way 383032616 from track to unclassified road (without even adding the unpaved tag). The imagery shows this goes through a forest, is unpaved, and runs steeply through terraced rice paddies. Chris labelled it a track for a reason, and I humbly apologise if VDL visited and its been asphalted since … but go on, prove me wrong. We don’t change existing data they say … more meaningless words.

The way forward … well I cant make bulk changes, I dont have the skills, but we should start taking control of the Thai map back … lets downgrade the short roads as Peter suggests, lets convert FB residentials to unpaved alley, until we can verify differently… lets convert all FB import=yes residential roads in the countryside, to unpaved.

When I started complaining about FB, I was a lone voice … I see Tom & Peter now following suit. Frederick of the DWG is correct; we never gave them any clear indications of what was wrong and what to do. As Thai mappers, we should meet and at least see if we reach a unanimous way forward.

We can do this, if we want to … lets not stand by and do nothing. You know, maybe we don’t even need the DWG involvement. I’m sure they will be happy if we make our own corrections locally. I certainly don’t think speaking to FB is any good, given the poor responses to date … it OUR BABY I’m afraid.

Rgds, Russ.

Thank you Russ, I think we are on the same page, except for one thing: I don’t think the DWG want to get involved. Like I said, the ball is in our court and it’s up to us to come to a consensus and then deal with FB directly. At least that’s what I take from Frederik’s email.

Perhaps we are a bit hard on FB: they may not have the resources (or experience) to fix existing roads much. I think if we could get them to stop mapping private driveways and mark most small stuff as tracks then we have a more detailed map that is then relatively easy to improve upon.

I still believe that a mass edit is called for.

Please keep me in the loop with a CM meeting. Perhaps I can swing a trip North.

Regards,
Peter.

Not exactly; but it’s difficult to get involved when there aren’t specific problems being pointed out in a form that we can do anything with. For example, above Russ has said " ask VLD012 why he changed Way 383032616 from track to unclassified road (without even adding the unpaved tag). …". That’s a good question - why hasn’t anyone asked them yet? The actual changeset I suspect is https://www.openstreetmap.org/changeset/55696040 ; there are no comments in the discussion. In fact, no-one has commented on any of that user’s changesets at all: http://resultmaps.neis-one.org/osm-discussion-comments?uid=6271594 .

Without details of the mismapped features that make up the “Facebook problem” there’s little that’s actionable from a DWG perspective - we can’t say “the local community asked you to do X, you’ve done Y, and we know that because here’s another changeset discussion where the local community have said that you’ve done Y again”.

Best Regards,

Andy (from the DWG)

Andy, that process is not going to work. We need to keep the discussion in one place. That’s why I proposed above to stick one forum topic. If we try to discuss every changeset individually then we will get nowhere.

We are talking about multiple users, thousands of edits over more than a year. The main problem is roads having been created that should either not be there or should be tracks. We also see imagery offsets .

All of this data, 150,000 km of it, needs to be reviewed, individually. Only FB would have the resources to do that. I am willing to do my bit and train one or two of their mappers.

Failing that the only viable alternative is to revert all their edits, as some of us here consider the map now polluted.

I do believe that some of their contributions are valid and valuable.

Let’s see what happens at the CM meeting.

Regards,
Peter.

If you want to limit the discussion to people who happen to be following this subforum then fine, but you can’t then say “I don’t think the DWG want to get involved” (for completeness, it was me that created the forum thread https://forum.openstreetmap.org/viewtopic.php?id=57387 in the first place as a heads-up to local mappers).

There’s always going to be discussion about “quality vs quantity” in OSM (there are lots of ongoing discussions about that right now on various OSM lists and in forums and chats), but without specific examples of specific problems there’s not a lot we can do. Rants on the forum certainly give a feeling of the scale of the problem, but they’re not really actionable. Right now there are about a dozen actual reported issues in OSM in the DWG queue, where someone has said “X is wrong because Y”. DWG members can either deal with those in whatever free time presents itself, or they can trawl through forum (and mailing list, and help question, and chat) threads looking for potential issues. Please help us to help you.

Also, let’s not forget about the QA done by mappers around the world in places that aren’t local to that area. Examples include geometrical impossibilities, tag misspellings etc - see for example http://resultmaps.neis-one.org/osm-discussion-comments?uid=5314378 . Changeset discussion comments are very useful to people doing that as they often indicate whether a problem is already known about and whether it has been acknowledged by its author.

Let’s meet. I personally mailed to the mappers in Chiang Mai I know. Posting it here for the (likely) case I missed someone.

Thursday 11th October, from 11:30 onwards, Shewe cafe
https://www.openstreetmap.org/way/528011072

For planning, please let me know if you plan to attend so I can arrange with the shop.

@Peter: Skype?

Thanks Stephan, I can make it, as I have to be in Phayao on the 14th. Count me in.

As mentioned elsewhere, I’ll attend.

Hi folks, I think our meeting was worthwhile and it was really good to put faces to the user names. Stephan took copious notes, so I won’t put a summary here, in case he wants to do it.

I do want to share what I just found and commented on: https://www.openstreetmap.org/changeset/63105877

Feedback is welcome. Am I on the right track?

Kind regards,
Peter.

let’s have a short summary of our meeting yesterday in Chiang Mai.

While there are multiple issues observed with the way data is mapped by Facebook (like number of nodes on ways not justified with the geometry, e.g., sometimes too few, other times too many), there are issues with the tagging of roads like potentially tagging driveways as residential or agricultural tracks as residential. Or unconnected highways where different imagery could have indicated a connection. Also sometimes trusting too much the positional accuracy of the imagery layer without cross-checking the alignment to GPS tracks.

The most annoying problem had been identified as agricultural tracks being classified as unclassified (or residential), providing routing software a good alternative to the proper roads by saving a few meters of distance. This also happens in case of a missed turn when the routing tries to bring you back on track and suggests to cross through some fields to U-turn.

As we have no good statistic about how frequent these problems really are, we decided to largely extend the changeset comments on discovered problems instead of reporting the issues in private messages or simply fixing them.

We also agreed that once a way was verified by a community member on the ground we are removing the import=yes tagging on the element.

We are also going to cross-check a random sample of these highways to get some numbers regarding how frequent such problems are. I will provide some tolling for this.
I will also provide a style for JOSM to highlight Facebook roads.

For checking changeset discussions, see here an overview of the last comments for Thailand:
http://resultmaps.neis-one.org/osm-discussions?c=Thailand&d=30#7/15.117/101.649

To get all discussions involving a specific user, go to
http://hdyc.neis-one.org/

and search for the username. Click on the link to view the commented discussions, then remove the “&commented” part from the URL. Maybe someone knows a more direct way to come there by knowing only the username and not id.

It was also asked how to see where a specific user has edited. The hdyc site from above can give some hints already. For details, you could search in the data, see here for an example: https://overpass-turbo.eu/s/CGY

To visualize changes, have a look at this: https://overpass-api.de/achavi/

yes, detailed comments like this help. Maybe not immediately fixing the things makes it easier to see what was actually done wrong.

Regarding way 629622938 intersecting with 2006:
After reverting your move of the node 5944737115 in changeset 63441790 it is not looking this bad. Facebook might have used the existing node instead of creating a new one but not that bad that I would have had complained.

Other points you had mentioned are correct. I added some more comments in that area…

Stephan

Sorry, you lost me here. This is what it looked like before I touched it:

https://bit.ly/2yBehdW

I think this is actually the most serious error of all.

(What is it with this forum s/w that it often displays part of the URL instead of the picture?)

I think you refer to a problem with the way attached to node 5944737115 intersecting the highway 2006, also making two right-angle curves in way 2006. Are you? If this is what you actually describe as a problem, then it was your move of the node to create the problem. My screenshot above shows the geometry after I revert you node move. Probably you accidentally moved the node when trying to pan the map.

Here is the history with a visualization of you moving the node four meters:

The current state of the map no longer shows this as I merged the road junction into the already existing node next to it.

I suspect that this node got moved with the whole area, when I adjusted the offset. I didn’t move the main road, as it was already correct and of course I have to click each node separately and if two are very close together then one of them will not get selected, as it’s not visible. I checked the history of both nodes and I was sure I my name didn’t feature in it, but who knows…

I’ll go and update my comment then.

Thanks for investigating.

Regards,
Peter.

Just a quick note to say thanks for the detail that people are adding in changeset discussion comments - that’s exactly what’s needed.

As promised on our meeting, here is a small map style for JSOM to highlight elements having “import=yes”, so in the majority of cases coming from Facebook.

The location is probably not forever, but I can leave the file there for the next few months, so you can directly add the following URL to JSOM as a map paint style.

https://downloads.osm-tools.org/FB_edits.mapcss

Thank you very much for the map paint style Stephan — it highlights the imports very well.

I visited a small town today, Ban Pin North, and didn’t really intend to do any major editing there but did want to add the name of the town to OSM. When I looked at the area in question using your paint style, the imports showed up very well. While there were no gross errors caused by the FB operator, my assessment is that many ways were carelessly drawn. The work, done about a year ago by RVR009, was generally not anything I would have wanted to have associated with my name.

I did not bother to write a changeset comment because my assessment was rather general and the work was done a year ago. Using GPS traces I adjusted the offset of the DigitalGlobe Standard imagery, and aligned ways (not all of which were created by RVR009), made a few obvious connections and removed the “import=yes” tag from the ways I checked.

I uploaded the imagery offset I used into the Imagery Offset Database as Ban Pin North.

Dave

I actually used to be indifferent to armchair mapping and imports, but recent encounters changed my mind.

Sigh

I really feel you now @Russ McD, @Tom Layo; one won’t see the reality in accuracy of these entered data unless one actually “live there”. Even the normal “imports” are usually based on data observed from on-the-ground. These corporate players seem to blatantly ignore the fact that fixing a map full of faulty data is much harder than adding correct data to a blank slate.


As a local resident in Bangkok, I had a recent run-in with corporate-based armchair mappers. In my case it’s Grab ridesharing company.

Their amount of changsets are much smaller than ones made by Facebook et al (around the magnitude of ~100 changesets); and as far as I know, they are 100% human-based team, not some nameless AI robots.

I began to notice the existence of their campaign, not because of their announcements, but due to the “obvious wrongs” that have been recently appearing in the area that I personally frequent to.

So I got curious, and began to do a proper on-the-ground verification of their edits: the result is not pretty.

With only just few samples, verified against ground observation, I found:

  • Overwrite of ground-surveyed data; broken local consistency.

  • Mis-upgrading of road levels (e.g. existing footway being upgraded to “service” despite the actual uses).

  • Mis-tagging of newly-added “roads” (e.g. new footway being tagged “service”); disastrous for routing.

  • Nonsensical objects tagging (e.g. a temple being re-tagged as road); ruined rendering.

Most of these “oops” are simply invisble from the satellite imagery, but glaringly obvious when you actually visit the place. Sifting through changsets and visiting places in their edits to fix them are like going down the rabbit hole, there’s no end of errors in sight.

And of course, I chided them directly in changset comments whenever I found these errors. But there are so many changesets that these will take several years to verify, in downtown Bangkok alone. Doing these erodes my sanity.

To date, I have verified just around < 0.5% of the edits done by their campaign. I estimate the count of errors in whole are to be many, many more. My current “verdict” is along the line of @Russ McD: the affected areas are now tainted; these can’t be trusted by satnav users anymore.

These are all I can say about corporate armchair mapping, from a very small subset that I actually went into the field to verify; now scale it to the entire rural Thailand.


Speaking of the scale of this “import” problem: don’t anyone realize that at the moment, amount of OSM edits from Facebook went through the roof and is now ranked as #1 in Thailand?

This is not a problem of AI; it is a problem of industrial-scale armchair mapping with zero ground-based verification. They are leaving their mess for us unpaid contributors to fix, case-by-case, for a several lifetime.

I condemn Facebook for this massive scale of data vandalism, and OSMF Data Working Group for their inaction.

With these, while I won’t quit contributing to my area in Bangkok, I don’t have the heart to recommend OSM for general uses in Thailand (and much less with satnav use) to anyone, anymore.

The damages are done. As a local user, my position on this is a firm: speedy deletion.

Erm - has anyone ever reported problems in Thailand with GRAB or Global Logic (their contractor) to the DWG ? I’m not seeing any tickets, but there may have been informal contact. I’m aware of reported problems in Indonesia (see e.g. https://www.openstreetmap.org/user_blocks/1882 and https://www.openstreetmap.org/user/harrymahar/diary/43671 from earlier this year), but nothing from Thailand.

For completeness - the list of GRAB contractors that I am aware of is at https://github.com/GRABOSM/Grab-Data/blob/master/Grab%20Data%20Team .

If there are current issues that need investigating with GRAB users please let us know. To be useful what we need are links to problem changesets and a description of what the problem is. Also if possible, with a changeset discussion history that shows that you reported the problem to them and they did nothing about it.

Please keep this thread strictly about Facebook edits. Please comment in detail at changesets what they did wrong. refer to the OSM id of the elements in question. Maybe even take a screenshot (snipping tool in windows). postimages.org allows hosting of them.

This is the task for Grab:
https://forum.openstreetmap.org/viewtopic.php?id=64075

Also comment here on changesets. As up to now we have very little to no complaints about them (they are not that active around Chiang Mai), please refer to commented changesets there as well.