New to NA so bear with me. I have several long routes created with Linear Referencing toolset for which I want to extract driving directions. The routes already exist so there is no sense in using Find Route tools or online routing services for directions, plus they are by no means least distance/cost/time/traffic-sensitive. For analysis purposes the only pertinent evaluators would be distances, road classes and oneways, in addition to producing these directions. Curious if the best course of action is to use NA tools. Will NA even support route created by Linear Ref tools? Any info is greatly appreciated!
Unfortunately, you won't be able to produce directions directly from your linear referenced routes. Directions draw from data in a network dataset and are produced from a route solved using one of the Network Analyst tools. The Directions need to know network-specific information, such as the streets that were used by the route, the street names, how long it takes to travel across the streets, etc. The linear referencing tools don't have any connection to the network dataset, so none of this information is available to draw from.
What's the overall goal of your project? If you describe it in more detail, maybe I can help you find another solution.
Thanks Melinda. I'm creating a set of long distance bike route maps that require simple block text descriptions for turn directions, including distances...."1.5 mi Turn right onto Main St, 5.5 mi Turn left onto Broadway, etc".
The linear referenced routes were created to generate other derivative map elements like elevation profile graphics, GPS tracks, and to calculate distances between turns. The missing data for these routes are the streets traversed, so I realized I would have to add this using some method. Creating a simple network dataset seemed like a possible good option. In the past we've used the Find Route tool and NA route services to recreate these routes and pull directions, but it was very tedious placing numerous stops along routes that are thousands of miles long.
Getting back to part of my initial inquiry, does NA support these linear referenced routes as inputs? Do I need to add the streets attribution before inputting?
No, the Network Analyst tools all require points as input. Perhaps you could pull out all the vertices from your linear referenced routes and use those as input stops for an NA Route. Solve the route and generate directions. If you create your own simple network dataset, you would have to set up Directions in the network dataset manually. It might be easier to continue using the Find Route tool.
Wow, really? That process (Find Route) is very slow and cumbersome for my purposes. Plus, the routes themselves are already resolved. Because these are bike routes, they often do not fit standard route calculation criteria (not least distance/time/cost and can often follow trails instead of roads), so they have to be forced a certain way. Hence all the stops.
Perhaps this is more general forum discussion, but do you know of any other toolset I can use to produce these directions from the linear referenced route (after adding streets data)? I figured technology was now there.....
Sorry, I meant that it might be easier to add your LR route vertices to the Find Route tool than it would be to set up a network dataset from your own data with good quality directions. It takes some effort to get Directions set up properly from your own data.
The reason I suggested grabbing the LR route vertices and using those as stops is that they are probably close enough together so that the Network Analyst Route solver would produce the same result as the original line you started with (rather than some different, optimized route). You can convert the LR lines to a point feature class of vertices using the Feature Vertices to Points tool (http://resources.arcgis.com/en/help/main/10.2/index.html#//00170000003p000000) and use those points as inputs to NA Route. If you are comfortable with python, you could automate the whole thing.
I do not know of any other tools that would allow you to generate directions from a linear referenced route. The directions engine works specifically with the Network Analyst tools and relies upon the information contained in a network dataset, which has no connection to the linear referencing results.
OK, now I understand. I'm curious about the directions that will be output if the LR vertices are used as stops. Like you mentioned, this LR route is pretty hi res and heavy with vertices. Will the resulting direction be showing a text line for each stop, at 0.001 mi intervals (the resolution of the route line)? If so, that would need to be cleaned up considerably. It seems to me another point dataset of turns would need to be input to determine the direction "breaks" that indicate turning from one road onto another. These turn points would seem to make better stops than the converted vertices, but maybe my thinking is incorrect. I should probably just test....
Yes, the resulting direction WOULD have a line saying "Arrive at Stop X", so it would get kind of messy. You could probably post-process that out automatically. The turns are included regardless. If you have to turn from one road to another in order to get from A to B, it will tell you "Drive X miles along Road P." "Make a left at Q street." "Drive Y miles along Road Q." "Arrive at Stop B".