Incorrect highways marked on Google Maps

[Roadsguy]

http://goo.gl/maps/RkvAz

[Central Avenue]

http://goo.gl/maps/RkvAz

AFAICT that's just Google's dumbass practice of not posting any sort of banners or green shields on Business routes.

Stupid, but not necessarily incorrect (you'll notice that if you get directions, it does indeed refer to it as "US 13 BUS").

[Roadsguy]

No, I mean the entire bypass isn't shown as a freeway.

[deathtopumpkins]

Google's completely screwed that up everywhere. There are countless non-freeways shown in orange, and countless freeways shown in yellow.

Here's an example: http://goo.gl/maps/fhYAB Binghamton, NY
-NY 201 on the left is a full freeway all the way from NY 17/I-86 down to NY 434, yet Google shows it as yellow.
-NY 363/7 on the right, however, has a signalized intersection just south of NY 17/I-86, and another one about a mile north of it, yet Google shows it as orange.

[empirestate]

Google's completely screwed that up everywhere. There are countless non-freeways shown in orange, and countless freeways shown in yellow.

There is no map legend for Google that I'm aware of...is it known whether the symbols do in fact represent freeways vs. non-freeways? I mean that would make sense, but...

[Central Avenue]

No, I mean the entire bypass isn't shown as a freeway.

Pfft, I didn't even zoom in far enough to notice those were grade-separated interchanges. Derp! 

[vdeane]

Google's completely screwed that up everywhere. There are countless non-freeways shown in orange, and countless freeways shown in yellow.

There is no map legend for Google that I'm aware of...is it known whether the symbols do in fact represent freeways vs. non-freeways? I mean that would make sense, but...

It's assumed so in the US and Canada, though in places like Mexico they're truly random.  The Mexican data has no means of differentiating freeways and non-freeways.

[Roadsguy]

Google has "expressway" and "freeway." The same color is used for both, though, which is dumb and confusing.

[vtk]

Google has "expressway" and "freeway." The same color is used for both, though, which is dumb and confusing.

I thought they just had "expressway", which was meant to be used for what MUTCD calls freeways, and more recently was applied also to what MUTCD calls expressways due to Map Maker contributors and moderators being dumb.

[Some_Person]

Google has "expressway" and "freeway." The same color is used for both, though, which is dumb and confusing.

I thought they just had "expressway", which was meant to be used for what MUTCD calls freeways, and more recently was applied also to what MUTCD calls expressways due to Map Maker contributors and moderators being dumb.

If Map Maker was more similar to OSMs editing system then Google would be the best mapping site there is. Or at least, they could just have everyone on this forum given editing privileges. Half the moderators on Map Maker know barely as much as we all do. Plus, I don't think 3 colors is enough to define all the different roads on Google Maps

[empirestate]

Plus, I don't think 3 colors is enough to define all the different roads on Google Maps

One principle I've always struggled with from a cartographic standpoint is: should the symbol used indicate a road's class (US, state, Interstate...) or its function (arterial, freeway, expressway and similar descriptors) or its physical characteristics (paved, unimproved, multi-lane...) or what?

Some maps, like the current USGS digital topos, use different symbols for Interstate, US, and state highways. Besides not always giving much of an idea what a road's really like (many state highways are freeways and of much higher quality than some backcountry two-lane US route), you could argue it's redundant because the shield symbol already tells you that.

Probably most maps try to depict roads according to their importance: primary, secondary, tertiary...or sometimes arterial, collector, local and so on. Freeways are pretty universally set off from other roads visually. But as we see at OSM, agreeing on a consistent and widely-applied symbology is harder than you'd think, partly because importance is relative. If a two-lane asphalt (or even gravel) highway is the only road for dozens of miles around, you'd say it's pretty darn primary. But a physically identical road in a suburban area would hardly register as important at all.

And how do you treat a road that every inch a freeway, except for one isolated place where maybe there's a grade crossing? You wouldn't downgrade the symbology for the whole road, but you'd also want to accurately convey the anomaly.

[vtk]

Some maps, like the current USGS digital topos, use different symbols for Interstate, US, and state highways.

That annoys the crap out of me, particularly when older versions of the map differentiated based on other characteristics.  USGS Topos and DeLorme Atlas & Gazetteers are both like that.  The problem is a switch to a digital roads dataset that knows nothing of a road's characteristics, but only its designation.  The new USGS topos actually have a roads layer derived from a commercial data provider, and that layer retains copyright status as a result.

[Roadsguy]

From my brief time in Map Maker, before I gave up on it, I could tell:

Orange:
-Freeway
-Expressway

Yellow:
-National Highway (pretty much only US routes...)
-Major Artery
-Minor Artery

White:
-I forget if it was "minor road" or whatever...
-Terminal road (slightly narrower, usually used for no-outlets)
-Some other really narrow thing

Ramps were assigned by "segment usage," not "priority." I still haven't quite figured out how to make a ramp into a non-ramp. It never seemed to work.

[Some_Person]

Ramps were assigned by "segment usage," not "priority." I still haven't quite figured out how to make a ramp into a non-ramp. It never seemed to work.

I'm glad I'm not the only one who can't figure that out like for example, here: http://goo.gl/maps/zThVX where PA 309 branches off I-78, those shouldn't be ramps, as it's the entire freeway that comes off the interstate, so it should be more like this example: http://goo.gl/maps/E0Qwk

[Bickendan]

This is why I much, much preferred the Thomas Brothers symbology.
They broke it down by road type -- Freeways and their ramps (thick red, thin red)
Highways/west-coast expressways (with at grade crossings) (thick tan, thin tan)
Arteries and their ramps (if they have interchanges) (thick green, thin green)
Collectors/secondary arteries (bold black line)
Local streets (thin grey line)
Unpaved (dashed grey line)

What's so difficult about adopting a theme like this for the online maps? Google maps would be excellent if it differentiated between road classes, so errors like OR 224 being shown as a freeway wouldn't exist (and it falls under the freeway category because its name is the Milwaukie Expressway).
OSM does a much better job, but it's not great either, because highways will shift colors for sometimes indiscernible reasons -- US 26 through Portland's a great example. It's a red line through Southeast Portland, shifts to tan east of I-205 all the way to the junction with Burnside Rd and the Mt Hood Highway, where it becomes green until the city limits of Sandy, where it goes red again. Why? Is it denoting maintenance divisions? I couldn't care less about that; I'd refer to an ODOT map if I needed that information. Is it a lane distribution? Then the portion through downtown Gresham is wrong, as it's four lanes, like the portion between the river and I-205. Is it a speed limit indication? Again, then it's wrong: From memory, Powell is 40 over the Ross Island Bridge, 35 from OR 99E all the way to about 122nd Ave, 40 from there to 182nd, back to 35 to Eastman, 30 to the Mt Hood, 55 to Sandy and down to 35 in Sandy itself.
Since non-highways also share the same colors, it muddies the water even further. Ok, green lines are the big, non-freeway highways maybe. Wait. N Greeley Ave and N Going St are green as well.

It's this type of thing that keeps a map from being excellent -- it's either simplifying too much and then falling prey to an archaic editing system with a bad review process (Google) or it's conveying too much information which would better fit a foreign system (OSM, which I believe fits the European road classification model better than the American one).

[empirestate]

This is why I much, much preferred the Thomas Brothers symbology.
They broke it down by road type -- Freeways and their ramps (thick red, thin red)
Highways/west-coast expressways (with at grade crossings) (thick tan, thin tan)
Arteries and their ramps (if they have interchanges) (thick green, thin green)
Collectors/secondary arteries (bold black line)
Local streets (thin grey line)
Unpaved (dashed grey line)

What's so difficult about adopting a theme like this for the online maps?

Couple reasons occur to me...

-This symbology works great in Thomas Bros. territory, but how do you apply it around the world? How would you show the Dalton Highway, for example? As a dashed gray line for unpaved road, or a thick green one for major artery? What about other countries where almost none of even the most major roads are improved?

-Functional classification isn't consistent with route designation in the U.S., but you still want to show some correlation for those actually trying to follow a route through an area. There should ideally be a symbological way to show the path of a US highway, for example, along different city streets that aren't otherwise physically distinguishable.

-Do you depict a road according to its design, or what traffic it actually handles? If a small road is choked with traffic, does it get the arterial symbol? Or does it get the minor road symbol, to emphasize that it isn't at all the right level of road for the traffic it carries?

[Bickendan]

This is why I much, much preferred the Thomas Brothers symbology.
They broke it down by road type -- Freeways and their ramps (thick red, thin red)
Highways/west-coast expressways (with at grade crossings) (thick tan, thin tan)
Arteries and their ramps (if they have interchanges) (thick green, thin green)
Collectors/secondary arteries (bold black line)
Local streets (thin grey line)
Unpaved (dashed grey line)

What's so difficult about adopting a theme like this for the online maps?

Couple reasons occur to me...

-This symbology works great in Thomas Bros. territory, but how do you apply it around the world? How would you show the Dalton Highway, for example? As a dashed gray line for unpaved road, or a thick green one for major artery? What about other countries where almost none of even the most major roads are improved?

If using a strict TB scheme, the Dalton (and similar highways in Yukon/NWT, like the Dempster) would be tan, to show that they are highways. That's how they handled OR 27 in the PNW atlas, which does have unpaved sections, if memory serves. A solution to note that it is unpaved would be in the highway label: DALTON HIGHWAY (UNIMPROVED [or] UNPAVED), with its route marker [AK 11].
What would be needed is an additional color for unpaved highways (unpaved arteries could use a dashed, uncased green line, for example, though that could clash with green marked areas, like forests, parks and golf courses). Another problem, one that I imagine RMN tried to solve, is colorblind people. Of course, RMN's symbology is fairly garish.

-Functional classification isn't consistent with route designation in the U.S., but you still want to show some correlation for those actually trying to follow a route through an area. There should ideally be a symbological way to show the path of a US highway, for example, along different city streets that aren't otherwise physically distinguishable.

If I'm reading this correctly, this goes along my point about US 26 through Portland and Gresham. That tan line from I-405 and SW Broadway all the way out through Gresham would tell me exactly where it goes (even TB [ok, RMN] screws this up, but that's really ODOT's fault for failing to precisely define how US 26 gets through downtown and then telling AASHTO: Market/Clay and Naito, per AASHTO? I-405, Broadway/5th/Sheridan, 3rd, Arthur, Kelly/Naito per ODOT?)

-Do you depict a road according to its design, or what traffic it actually handles? If a small road is choked with traffic, does it get the arterial symbol? Or does it get the minor road symbol, to emphasize that it isn't at all the right level of road for the traffic it carries?

Artery. SE Division St springs straight to mind. From 7th to 60th, it carries a lot more traffic than it was designed to; it's also a straight shot from Gresham to Portland for commuters, running from SE 3rd to 317th out past Gresham. I see that green line, I know to keep a bigger eye out for traffic when walking/riding a bike/etc.

[empirestate]

-Do you depict a road according to its design, or what traffic it actually handles? If a small road is choked with traffic, does it get the arterial symbol? Or does it get the minor road symbol, to emphasize that it isn't at all the right level of road for the traffic it carries?

Artery. SE Division St springs straight to mind. From 7th to 60th, it carries a lot more traffic than it was designed to; it's also a straight shot from Gresham to Portland for commuters, running from SE 3rd to 317th out past Gresham. I see that green line, I know to keep a bigger eye out for traffic when walking/riding a bike/etc.

OK, that's all fair...but what if I'm using the map to find the best route through, assuming that the map will show me the roads that I can expect to use without issues of congestion...or, if no such route exists, to show me that instead?

In other words, it comes down to the fact that maps should show what they're designed to show, and there's no catch-all answer for that. If I make a map explicitly of Interstate, US and state highway systems, then by golly I had better use symbology that reflects those distinctions! But if my map is supposed to be for effective motor travel, I probably should use symbols based on functional class instead. And you've mentioned walking, biking, etc.–maps can be separately designed for those uses as well.

The challenge, then, is what to do with online mapping, which ambitiously tries to be an all-in-one solution. Many, if not most, people may use them primarily for navigation, but there are other ways they are frequently used as well. As a result, different providers have different levels of success. I use Google for the most up-to-date aerial imagery, MapQuest for the most up-to-date map date, and Bing if I want, for example, to see all of the counties successfully labeled at a single zoom level. I use OSM, on the other hand, to look at areas where I want human interpretation to have been applied. Where it has been, it's most accurate of all, and where it hasn't, that's fairly evident from the map's appearance. Then again, I'm a much more advanced map user than the average Joe...

[kphoger]

what if I'm...assuming that the map will show me the roads that I can expect to use without issues of congestion...?

But who reads a map with that assumption in mind?  Aside from the freeway/surface route and single/dual carriageway distinctions, what congestion issues does one really expect a map to convey?  Ratio of number of lanes to traffic counts?  Stoplight timings?  Speed limits?

Yeah, that kind of stuff might be useful on a map, but I don't think people (OK, more than seven people) really think that's...

what they're designed to show

[empirestate]

what if I'm...assuming that the map will show me the roads that I can expect to use without issues of congestion...?

But who reads a map with that assumption in mind?

Why, travelers using the map, of course.

Aside from the freeway/surface route and single/dual carriageway distinctions, what congestion issues does one really expect a map to convey?  Ratio of number of lanes to traffic counts?  Stoplight timings?  Speed limits?

No, think more broadly: mountains, cities, that sort of thing. If I'm reading a map and I see a big fat line in a bold color running through an otherwise blank area, I'll assume my passage there will be relatively easy. Now, if a dot or a blob with a city name appears along that line, I now assume some kind of settlement. Perhaps the road I'm on barrels through that town with essentially no interruption, but perhaps instead it gets funneled down a single street that's far too narrow. Yet it's the only way through, and many people pass through this town, so it's completely choked with traffic, and it may take me an hour or so to traverse this little town. I think it's a fair assumption that a map of reasonable detail will somehow show which of those cases is more likely.

So as for specific highway symbology, would you retain the bold, fat line symbol on that road as it passes through that town, to show that it carries all of the traffic through–i.e., is the artery? Or do you switch to a skinny little line to show that the way through town is along a skinny little street, hoping the reader will infer that the trip from big fat line through skinny line and back to big fat line will add some time to the journey?

[NE2]

But if my map is supposed to be for effective motor travel, I probably should use symbols based on functional class instead.

When editing OSM I try to use a sort of polished HFCS-style system. HFCS itself is a bit too messy to use on a finished map; classifications are not always moved after major realignments, and sometimes begin or end at arbitrary places.

[kphoger]

If I'm reading a map and I see a big fat line in a bold color running through an otherwise blank area, I'll assume my passage there will be relatively easy. Now, if a dot or a blob with a city name appears along that line, I now assume some kind of settlement. Perhaps the road I'm on barrels through that town with essentially no interruption, but perhaps instead it gets funneled down a single street that's far too narrow. Yet it's the only way through, and many people pass through this town, so it's completely choked with traffic, and it may take me an hour or so to traverse this little town. I think it's a fair assumption that a map of reasonable detail will somehow show which of those cases is more likely.

So as for specific highway symbology, would you retain the bold, fat line symbol on that road as it passes through that town, to show that it carries all of the traffic through—i.e., is the artery? Or do you switch to a skinny little line to show that the way through town is along a skinny little street, hoping the reader will infer that the trip from big fat line through skinny line and back to big fat line will add some time to the journey?

Neither, since the black dot covers the town and obliterates the red line in that location.  Fat lines should denote trunk highways, and skinny ones should denote feeder highways.

[empirestate]

So as for specific highway symbology, would you retain the bold, fat line symbol on that road as it passes through that town, to show that it carries all of the traffic through—i.e., is the artery? Or do you switch to a skinny little line to show that the way through town is along a skinny little street, hoping the reader will infer that the trip from big fat line through skinny line and back to big fat line will add some time to the journey?

Neither, since the black dot covers the town and obliterates the red line in that location.  Fat lines should denote trunk highways, and skinny ones should denote feeder highways.

OK, fine...  Suppose it's a really long stretch of narrow highway, like through a mountain range, if that helps.

Or don't suppose anything...bottom line is, do the classes "trunk", "arterial" and so forth refer to the traffic a highway carries, or that which it's designed to carry, or capable of carrying, even though the actual amount is much more or less? I'm speaking from a cartographic standpoint, not an engineering one.

[NE2]

I'd say it's more recursive than that. If you have two parallel routes connecting nearby points, and one is clearly a better route to take, that one should be a higher clasification. Classifications should end at logical points (usually towns or major intersections). So if a six-lane highway becomes a two-lane main street, it should keep the same classification *unless there is a bypass*. This, of course, ignores classifications based on physical characteristics (OSM's trunk, sometimes; Google's expressway; paper maps' divided highway). My personal rendering preference would be the way I've seen some paper maps do it: color indicates importance and thickness/number of lines indicates width. This is trivial at low zooms, where one doesn't have to deal with one-way pairs (the occasional super-wide median looks a bit ugly, but whatever).

So for the example of a four-lane going to two through a mountain range:
       ^ ^
===-----===
        ^

[Bickendan]

Is that I-93 in ASCII?