Hacker News Reader

Comment by liquidgecka | Hacker News Reader

Google keeps adding the irrigation access road next to me to maps. This is both dangerous as its totally unprotected and a car could easily drive into the canal, and its private, only the property owners and irrigation district are allowed to use it. I have reported it something like 10 times already and every tie it will be removed for a day or two and then instantly pop back into the map. I don't get how Google even thinks this is a road, or what data it is using to discover it. Its nowhere on any map that I know of!

View on HN

It's pretty likely that the road data source (e.g. the local government) marks the road without any restrictions. Data errors are shockingly common with municipal sources. Even if that's not the case and the source correctly marks the restrictions, maps will still route onto it for destinations "along" the road.

This is what I believe to be the cause behind most of the anecdotal stories in the comments here. Google Maps gets bombed with fake edits by people who don't want others using "their" road, so in the event of conflicts map providers have to defer to official sources.

When I was at Google I tried to track down someone on the Maps team open to adding a "Jeep Mode" that would let you opt-in to routing via roads that were unavailable and telling you why (washed out, poorly maintained, etc). Nobody seemed to like the idea.

Many people who haven't had the misfortune of working with geospatial data underestimate how unbelievably hard is to make a good map, let alone a global one.

It's quite fun trying to work out where an actual walking track is based on a printed topographic map (has good coverage of advertised official tracks but misses some), OSM (has nearly everything but tends to hallucinate non-existent tracks), the output of a local map-maker (compiles OSM and other sources but sometimes has the same track twice in slightly different places), and Google etc. (doesn't show anything a hatchback can't traverse, but has good satellite imagery that can show some tracks).

This weekend I'm going to go explore one of the OSM tracks and see where it's a total hallucination, old and overgrown, or real.

What do you mean by "hallucinate"? FYI most of the time it's a human drawing what they're seeing on the aerial imagery.

I know many instances in which someone mapped a track through a field which subsequently got plowed by the field owner, but otherwise I have no idea. Could you tell any specifics?

Imagine you draw a line following a trail on a map. Looks good, ship it! Now someone else loads that same line on their computer and it looks like a jagged mess because they are twice as zoomed in on the map as you were. Oops. What about the wide parts vs narrow parts of the trail? Do you represent that as two lines or just keep one down the middle? You send your trail to a friend who overlays it on their map that uses a different datum? How were the photos from the airplane assembled, and how did they represent a round object on a flat image? Did they have a good GPS lock when they took the photos, or did they even bother to check? What if the line you created is to be used 100 years in the future when the center of mass of the earth or magnetic north has shifted, how does your coordinate system account for that?

I know just about enough GIS stuff to fit on a few index cards, but I quickly realized it is way more complex than how simple Google Maps makes it out to be.

Sorry, hallucinate in the sense that the map says there's a track but there isn't anything there, whether or not it was drawn by a human. I meant it as an allusion to the map tripping balls but forgot that the overlap with AI terms would confuse HN.

With OSM I think a lot of walking tracks are from GPS traces uploaded by volunteers, so maybe sometimes stuff sneaks in. Or maybe the map shows old tracks that have since been erased by rangers. Or maybe there's a nearby track that has been marked down in the wrong place.

Is this because of the data quality and difficulties with merging across CRSes or otherwise slightly differently aligned datasets?

I dabble in geospatial data for my work in air quality modelling, accuracy for me means being within approximately 10-50 meters. Even under those very loose constraints I find the data handed to me by clients is often not good enough. Incomplete AutoCad dxfs, PDFs, and dubious SHPs all make the work much harder than it needs to be.

It's everything. The problem domain for Google maps is effectively "everything on Earth". I haven't worked with gmaps specifically and can't claim specific expertise there beyond some social conversations with people on the team, but some fun things I've encountered include:

* Roads that have no physical delineations (e.g. an entire valley is the road).

* Landscape features can change rapidly (e.g. forest fires rendering roads unusable on a minute-by-minute basis).

* Landscape features can change slowly (e.g. erosion, continental drift)

* Landscape features can change slowly and rapidly (e.g. features on glaciers).

* Data that's flat out wrong for political reasons (e.g. disputed territories, GCJ-02, hidden "things").

* Data that's flat out wrong for technological reasons (e.g. GPS jamming).

* Things that are named in languages without written forms.

* Things that have official names other people consider offensive (e.g. Piestewa peak).

When you make a map, all of these and more become your problem. Every edge case anywhere in the world is potentially something you have to worry about.

In addition to what the other commenter said, disparate sources with incompatible schema, often updated at a glacial pace are a big problem and possibly a root cause of the issue in the article.

Thanks for trying. I don’t use Google Maps, but if they had a Jeep Mode I’d absolutely use it.

Of course, every now and then Google does send someone down the path less travelled by. One day while exploring a remote area and reaching the point where flooding prevented me from going farther, a couple pulled up in some massive SUV, having driven down a road that must have given them pause.

Driving from Oxford to Swanage via Poole, looking at the satnav, I noticed that we were passing near to the village of Middle Wallop, and commented on it to my driver (bec ause it's a cool name).

We started getting directed down stranger, narrower roads, until we came to a muddy farm track in a tiny village, at which point the satnav announced "You have arrived". We were only halfway to Poole.

I must have momentarily touched the satnav screen, and inadvertently reset the destination to Middle Wallop.

On the same trip, on arriving at Poole, we should have taken the ringroad; but the driver ignored my navigation directions, and we ended up on a highway into Poole centre. I told the driver to just do a U-turn, and return to the ringroad; then the satnav announced "Board the ferry". We looked at one another, simultaneously saying "What ferry?" Then we looked up, and there was a car-ferry, loading up. So off we sailed to Swanage. Better than taking the ringroad.

I figured it was a 'smarter' system based on image recognition of satellite data. Or perhaps based on location data of people traveling.

I guess this option is much simpler. And it is so frustrating, because it is almost easy to fix, if only you knew the real data source. Especially if it is a local government, they should be quite responsive.

Doesn't Street View mean they have images showing where the gates and private road signs are?

Our local highway district actually makes this data available for free so I checked, its not in there. I also called them and asked about it and the lawyer seemed annoyed that google was adding county easements as public roads too since they alone should have access. So its not coming from the county at least. Its also marked roads where farm tracks are that are not in any way associated with public roads or access. I suspect that there is something looking at the photos taken by aircraft and thinking "that looks like a road!" or something because nothing else would make sense. Why mark something as a road that is just a space between two crops that is driven by tractors?

Also worth noting, its not on OOSM or Bing.. Only Google. So the source is something that only Google is using.

Yeah, they treat the so-called authoritative data as potentially incomplete and supplement it with data derived from imagery. The intent is that the authoritative data will override supplemental data, but if it's simply not there they'll use whatever else they can see.

> I don't get how Google even thinks this is a road, or what data it is using to discover it. Its nowhere on any map that I know of!

I suspect that someone holding an Android phone simply drives the path.

If you're Google, which will you trust more:

1) A bunch of sources of varying correctness which may/may not have been tampered with.

2) Actual GPS coordinates from your OS saying "Hey, somebody drove here."

I don't think gmaps actually does that. We have a 20 meter stretch of road near our house that connects two other roads. Google, for at least the last five years, has refused to believe that this road can be traveled, having everyone take a detour. Hundreds of cars are taking this non-existing road every day, but Google still hasn't taken the hint. (Probably preventing quite a bit of through-traffic in our neighborhood, so we're actually rather happy about this. :-))

20 Meters is quite short, so assuming a certain speed, there might only be a few GPS probes for each pass, that might also look as if they were outliers.

But how does that explain everybody getting from a to b without taking the two minute detour?

While I work at a company that, among other things, builds a navigation system (with a twist of doing load balancing on streets; called NUNAV Navigation), I'm not really an expert in this topic. From working with GIS and GPS quite a bit, I know though, that a lot of things, while they are working very well, are not as exact as they look (e.g. bad GPS signal, especially when driving with low speed, map matching, etc.)

Maybe someone else can comment in more detail.

I don't know if Google Maps pays much attention to Android users frequenting a particular path. Where I live you need to take a particular route else you end up at closed gates. Despite hundreds of people using this route every day for years Google will still try to take you via a different route that takes you to a closed gate. Other mapping software gets it right but we have to provide specific route instructions in case anyone uses Google and ends up at a dead end. I have noticed this all over actually, with Google trying to take short cuts that don't work vs following well-established routes.

I thought that too, but the irrigation district locked a gate when they finished up in October and yet it is still popping up. So if they are using phone traversals then its all foot traffic and that is more terrifying than it being detected via satellite images!

Clearly needs 'hazard documentation' features (legal, safety, otherwise) to prevent roads that shouldn't be used from being used.

They do that to us too. Google wanted me to take a detour over an access bridge that allowed a farmer to cross a ditch (man made creeks in the Missouri Bootheel).

There's a semi-maintained road that goes through a farm field between my parents house and my childhood best friend's house. Partway through there's a creek and the road just goes through the creek because it's usually dry or shallow enough you can drive through it with a truck. It gets pretty deep and wide in the spring though. Google Maps suggests that route. Also, I'm pretty sure it's a private road and likely not passable in the winter because it probably doesn't get plowed.

Does it exist on OpenStreetMaps and has it been marked restricted there?

I have a small suspicion that Google is cross-referencing itself with OSM and some changes stick if OSM corroborates. But it won't ever agree to the road not existing (because it does exist).

I think you can fix it from Open Street Map. They use OSM a lot.

Wonder if the people who do use it want it back.

The people that use it (legally) are the irrigation folks. They use it every day and the property owners can not block access. The private property owners are all commercial developers that have held onto the empty land for a few decades and seemingly have forgotten that they own it given that they never show up in the first place. I can't think of anybody else that would want that to be on maps at all.

Reply on news.ycombinator.com