Finding distance (in time) from nearest point out of 4 along road network using ArcGIS Desktop?

Question

I have four points representing harbors along the coast. I also have a roads layer that has info on each road segment such as length, speed limit, and travel time.

I calculated travel time with [(length in miles) / (speed limit in mph)] * 60 for the time it takes to traverse each road segment in minutes.

What I want to do now is: Work out from each harbor point along the road networks to get the total travel time to the nearest harbor for each road segment. All the tools I've found that do something similar to this require both a start and end point.

Is there a way to do this without turning each road segment into a point itself?

To better visualize my end goal, I essentially want a sort of "Euclidean Distance" effect, except it needs to stay along the roads layer and will total in time rather than meters/miles/etc.

I am using ArcGIS Desktop 10.0.

Answer 1

Using Network Analyst extension, you could:

1 - Use OD Matrix analysis, with the 4 harbor points, using travel time as cost, to calculate "nearest" facilities from each other;

2 - Select the 3 "nearest" pairs of harbors;

3 - For each of the pairs, use route analysis to create best route between them.

4 - Dissolve resulting lines summing the time;

Note: If your road network has oneways, going from harbor 1 to harbor 2, might be different from going from harbor 2 to harbor 1

Answer 2

You might take a look at the Closest Facility solver of Network Analyst. Then each harbour will become a facility and three others - incidents. Then you would do the opposite for each harbour - another one will become a incident, others - facilities and so on. I doubt you need ModelBuilder or Python here - since you have just four points. Then you can aggregate the travel time taken for every road segment along the routes between harbour points. I would suggest taking a look at an add-in which makes it possible to convert the result routes into a feature class (with edges, turns, junctions) which you may use for further geoprocessing.