Select to view content in your preferred language

GIS high density vehicle routing soloution

7079
14
06-19-2013 01:29 PM
lucasmostyn
Deactivated User
Hello

I've been asked to provide an optimized routing soloution for the rubbish and recycling trucks in my district. I do not have access to the Routesmart extension so I've had to work with the Network Analyst 'Route' or 'Vehicle Routing Soloution' options.

So far I've had most success with the 'Route' option by manually creating polygons of the service areas, then intersecting this data to all of the city street corner house numbers, and then using the zones field from my manually created polygons to organise the days collections. This is essentially a table based system of solving the problem.

The problems are really starting to pile up though. My routes are certainly not the most efficient. I've lost all of my truck capacity functions, and although I am capable of routing trucks from the depot to the route, to the recycling centre, back to the route and then onto the depot again, I have no way of doing the drop offs intelligently. That is to say I lost most of the benifit of optimizing a route.

Does anyone have a soloution or any ideas?

p.s I have used the city street corners to try and remove some of the density from the problem. It would be more ideal to be able to use all of the street addresses as i would be able to calculate capacity more effectively...

Any thoughts would be appreciated...

Regards,
Lucas
Tags (2)
0 Kudos
14 Replies
lucasmostyn
Deactivated User
Max orders are 600
0 Kudos
JaySandhu
Esri Regular Contributor
600 stops should be doable. Have you tried some other solver such as load them as stops in a route layer and solve. Do you get a path? If yes, you could turn on the find best order option and run again (this is similar to vrp) and see if it can solve. Other than that I cannot say what is going on preventing its completion.

Jay Sandhu
0 Kudos
lucasmostyn
Deactivated User
Hi Jay

Yes I have tried loading them into the the 'Route' solver, and yes it does generate a route. However, I cannot get an optimized route this way. If i check the box to optimize the route order then the process freezes when over 400 points are used...

If I do not check the optimize route box then I can solve over 10,000 points.... (but this is only presenting a route between the points in the attribute table without sorting them into a good sequence)

I think this process would work if I could find a way to organise the points into the correct sequence before running the 'Route' solver...
0 Kudos
JaySandhu
Esri Regular Contributor
When it "freezes", what is the memory status for the ArcMap process? How much RAM has it consumed?

Also, can you send me your points (export to shape file and zip OR save the route or vrp layer and zip and send) and I can run them locally to see if I can get a solution? jsandhu at esri.com

Jay Sandhu
0 Kudos
KarelKus
Emerging Contributor

Hello,

is there a solution for this problem, I am dealing with similar thing - Micro routing for recyclables collection, and I also have to use Network Analyst.

I have two zones for collection 816 and 802 (nodes) bins.

After clicking solve the algorithm runs well. For 816 nodes it takes an hour, but after that time I get a sort of  solution. I have outputs in route properties (TotalOrderServiceTime, TotalTravelTime...), but there is no route on the map. I can see only the sequence of orders.

In addition two error appears:

Error: "Failure while working with internal route problem instance"

Error: "A column was specified that does not exist."

I think I did sometning wrong with my network dataset or attributes but still cannot find out what it is.

Any suggestions?

Anyway the sequence of orders I can see is not the optimal one because some nodes are skipped and the truck returrns for them later. 

1. Is this caused by that the solver is not designed for such a task?

2. I started experimenting and this anomalous sequencing occures on smaller sets too. Now It seems to me that it depens on the number of nodes per route, but also on the exact area in the network and set of nodes that is chosen. Sometimes it helps to use a seedpoint sometimes not.

Is there some exact threshold how many nodes the solver should be capable to do without this sort of error?

I would need 255 nodes for each of my zones to solve it by placing one point along the mid point of an edge.  

0 Kudos