Page 8 of 57

Re: Road Types (USA) – comprehensive overhaul of drivable ro

Posted: Thu Mar 20, 2014 3:14 pm
by CBenson
Good point, that would justify keeping the exits and entrances as ramp type.

But that logic would also confirm the exits and entrances as "at-grade connectors" that meet the exception to be typed as ramps. I could certainly see some editors making the argument that would mean that the freeway definition of "no at-grade intersections" is not met as you are applying the rules of "at-grade connectors" to the exits and entrances. That argument would result in US-50 being typed as a major highway here.

The advantage I-75 has is that it is an interstate. Thus under the proposal it is a freeway and thus all entrances to and from the freeway are defined as ramps.

Re: Road Types (USA) – comprehensive overhaul of drivable ro

Posted: Thu Mar 20, 2014 4:17 pm
by CBenson
sketch wrote:There's a nuanced difference between the definitions here.

A freeway has no at-grade crossings
I like that way of putting it, but don't see "crossings" in any of the proposed pages.
sketch wrote:and no intersections (and no immediate access to adjacent property). There are no at-grade crossings on US-50 there. An "intersection" requires that the roads themselves intersect, not via ramps, but on their own. So, those aren't "intersections", and that's a freeway. There's nothing to stop a ramp from being at-grade.

For at-grade connectors, if we don't want to have to rely on the "signed, numbered exit" exception (which should be expanded to encompass all BGSes regardless of number), we can either (1) define it circularly, saying that every road providing entrance or exit along a freeway must be a ramp,
It is the circular definition that is causing me to worry. A ramp is a entrance or exit along a freeway (or something of a similar style on a highway) and a freeway only has entrances and exits via ramps.
sketch wrote:or (2) change the definition of "at-grade connector" slightly, to include only those at at-grade crossings or intersections. The latter (2) is logically more sound, but then what do we do with situations like this?
I was thinking about the RCUT under discussion with respect to the proposal to add the MUTI guidance to the general wiki. I agree that highways using RCUTs should be typed as highways not freeways. Although, in Maryland these roads using them seem to be classed as OPAs, so under the proposal they would be typed as major highways anyway.

I'm leaning toward just using the "at-grade crossing" language to define a freeway rather than trying to define what intersects with what. I would argue that there are no at-grade crossing on US-50, thus it is a freeway and the at-grade connector rules do not apply to entrances and exits of a freeway. I would argue that for the RCUT and your example the main road does include at-grade crossings thus even if classed an expressway it should be typed as a major highway. For the RCUT you can cross the highway and you do it without a bridge by making the right/U/right, thus it is an at-grade crossing. For connectors to a major highway where there is no bridge (grade separation), one should look to the at-grade connector guidance.

Re: Road Types (USA) – comprehensive overhaul of drivable ro

Posted: Thu Mar 20, 2014 5:26 pm
by CBenson
sketch wrote:"Crossing" to me means the road crosses from one side of the freeway to the other; i.e., it doesn't end/begin at the freeway. The reason "Intersection" should also be included is that it also covers roads that do end/begin at freeways — but only those that do at, well, intersections. Perhaps an "intersection" should be defined as "roughly perpendicular", where a road and another actually intersect, without the exclusive use of connectors or ramps.
Understood, but I guess I'm still not convinced that intersections that are not crossings need to be considered in typing a highway/freeway. Can't we simply define a freeway and a major highway by the roads that cross from one side to the other and let roads that do end/begin at freeways take care of themselves. If roads don't T in at traffic lights (which would disqualify the road as a freeway) can't the road still be a freeway if there is an intersection but no crossing. Should US-50 be a major highway because of this intersection? If so how much of US-50 should be typed as major highway?

Re: Road Types (USA) – comprehensive overhaul of drivable ro

Posted: Thu Mar 20, 2014 6:13 pm
by CBenson
sketch wrote:I don't envision anyone really having a problem with that being called a freeway any more than they might have had before.
I absolutely agree. I think the proposal is an improvement. Please take my comments as constructive criticism intended to make them better than the current guidance. The reason that this section came to mind is that there were questions regarding what this road should be typed under the current guidance. I was hoping to clarify the issue under the current proposal.

I'm not too worried about the the Baltimore Annapolis Rd intersection, there are always going to be the oddball exceptions. So maybe we don't need to anything else.

Re: Road Types (USA) – comprehensive overhaul of drivable ro

Posted: Sun Mar 23, 2014 7:39 pm
by CBenson
qwaletee wrote:There's already code to treat unnamed ramps as if they were named with the same name as the next named segment on the route.
This isn't a ramp property. This is true of any segment, isn't it?

Re: Road Types (USA) – comprehensive overhaul of drivable ro

Posted: Mon Mar 24, 2014 7:29 pm
by CBenson
I like both the font change and the chart.

Again the one thing that I would bring up is the signed/unsigned distinction. It's likely fine, but everything signed as an interstate does not have the functional class of interstate. Conversely, not everything that has a functional class of interstate is signed as an interstate.

For example, the entire Baltimore beltway is signed as I-695. However, the Francis Scott Key Bridge over the Patapsco River is not built to interstate standards and thus some of the beltway is officially designated as MD-695, given a functional class of "other freeway/expressway." It does remains fully access controlled, so there is no question that it should be freeway type, so this is for the most point a distinction that doesn't matter. But I'm not sure which block in the chart this section of road should fall.

On the other hand, US-50/John Hansen Hwy from about the Washington Beltway to about MD-70 at Annapolis is officially I-595. However, it is only signed as US-50 (with several other concurrencies in some stretches), there are no I-595 signs. It has a functional class of interstate. Again its a limited access highway, so there is no question that it should be freeway type.

I defy you to follow US-1, US-29 or US-50 through Washington, DC by following signs. There are a couple of issues, first the roads used for the east/north route may be different than for the west/south route (so a segment can fall in a different box in the chart depending on which direction you are going on it). Another issue is that the signage as far as I can tell is utterly random. Its not non-existent but there are large gaps in coverage and few signs tell you to turn. The signs seem to be simply informing you every once in a while that you happen to be on a US highway. There is no effort to enable you to follow the routes by following the signs. I'm sure that there must be some official designation somewhere regarding which streets these US routes follow through DC, but if you can't follow the signs should the roads really be bumped to major highway simply due to the official designation?

I recognize that these are edge cases, and that the guidance doesn't need to be comprehensive. But I bring them up for your consideration as to what they are worth.

Re: Road Types (USA) – comprehensive overhaul of drivable ro

Posted: Fri Mar 28, 2014 1:48 pm
by CBenson
Thortok2000 wrote:It should route off PURE traffic data in my opinion. If the traffic data shows that road is the one to take, it should take it. It shouldn't take very long at all before the 'major' roads naturally become part of the majority of routing in the area. If you were to label EVERY SINGLE ROAD a major highway, you'd probably get the absolute 'best' route every time. However, it's easier for the routing system to 'think' if it only has to consider and compare 3-4 options rather than 30-40. Still, I think comparing every possible route regardless of road specification and finding the true fastest route is what we would love Waze to do, although there's little that we can do as map editors to make that happen. So FC it is.
You can't possibly really mean that waze should check every possible route between two points. There have to be methods for pruning unreasonable routes from consideration. Taking the back roads across long distances between cities isn't going to be the fastest route and looking at every back road route would overwhelm the calculations. The question is how to define the back roads. Waze understands that editor decisions of road type should not be the only factor in this decision and is working to include their own speed and volume data to make better routing decisions.
Thortok2000 wrote:So the more major highways you have, the more routing improves along the major highways, but the less it works for lower classifications. Waze is like "I want to take a major highway and I'm going to unless there's a STRONG reason not to. If there is, I'm going to take only majors and minors unless there's a STRONG reason not to. If there is, I'm going to take only majors, minors and primaries, unless there's a STRONG reason not to." It's because of this 'STRONG' reason that it sticks to these routes like glue when shortcuts or shorter paths would actually be a bit faster.
I guess it comes down to what makes a STRONG reason. Almost every day waze routes me of off highways onto primary streets where there are highway alternatives. I am in an area that I suspect waze has a lot of data for. So I suspect that the effect of the road types is much diminished where waze has a lot of actual speed data to rely upon.

Re: traffic data only

Posted: Fri Mar 28, 2014 8:12 pm
by CBenson
dbraughlr wrote:I do not want routed through areas where I have to watch for children running into the street or trucks backing into loading docks.
I the case waze is counting: I do want to be routed through such areas.

Re: Road Types (USA) – comprehensive overhaul of drivable ro

Posted: Sat Mar 29, 2014 9:55 am
by CBenson
daknife wrote:(I also admit to a slight dislike for the Major type because the color used in the App is too close to the heavy traffic color occasionally making it hard to tell at a glance whether that's a Major hwy ahead or a major traffic jam that I need to avoid.)
I suspect very few users that don't use WME frequently use the map editors color scheme. The map editors scheme is almost by definition not optimized for the best user experience.

Re: Road Types (USA) – comprehensive overhaul of drivable ro

Posted: Sun Mar 30, 2014 3:56 am
by CBenson
Thortok2000 wrote:Not once has it ever offered me any other solution to avoid that N Church St traffic; it's always to take Academy instead. Never anything else. Ever. Ever ever ever ever ever ever ever ever.
So what happens if you ask for alternate routes? When I bring up your live map route I'm currently given this:
Wade.png
(163.26 KiB) Downloaded 802 times
Waze seems to consider going down Main St, but concludes that its over 10 minutes slower. I don't know why waze thinks its that much slower. But yes, it seems that waze isn't going to find Main as the faster route unless waze detects a major backup on both Church and Academy that results in those routes taking 10 minutes longer than usual to go the 2 miles.
Thortok2000 wrote:So that's my experience. The only main reason I can think of why my experience is so much different is because there isn't a strong amount of Wazers in this area. Which makes me wonder why a standard that doesn't fit for areas of low usage would be a good standard to apply nation-wide. =/
So I will just say for the record that my experience is very different. On my morning commute, which is longer than yours, I have documented waze's initial calculation as giving me over 2 dozen different routes and there are likely a couple dozen more routes that I haven't documented. Plus waze will also change my route during my drive. Granted there are many minor variations for a few major routes. But the point is waze does route over these minor variations for me.
The difference may be the higher usage of waze here. The difference may also be the server giving me different routes as I volunteer for various beta tests. The difference may also be that traffic frequently comes to a standstill around here, so maybe the traffic here is severe enough to overcome whatever stickiness you experience.