Snow Plow Routing

4141
2
06-27-2011 08:55 AM
Highlighted
New Contributor
We are a county with 504 square miles, a population of 36,000, and 405 miles of County Roads.  We have created maps for snow plowing but I am wondering if there is a way to improve our plow maps.  We have 14 route drivers. 

Factors that influence plowing are:  Number of passes required due to shoulder width, posted speed, other traffic, snowfall accumulation, intersections, and sight distrance (wind). 

Sometimes one part of the county will experience significantly different snowfall or wind.  Ideally we would like the 14 route drivers to complete their routes at the same time.

My questions are: 
1. Is it worth the cost FOR US to purchase Network analyst when there are so many variables? 
2. Even if we dismiss the variables and only account for mileage, number of rounds required due to shoulder width, and posted speed, will network analyst produce 14 logical routes?
3. Is there another tool we should consider instead?
Tags (2)
Reply
0 Kudos
2 Replies
Highlighted
Esri Regular Contributor
Christy,

Snow plow routing as you describe it is technically called a Chineese Postman Problem or Euler Tour where you have to visit every edge (not discrete locations) with a variety of constraints.

Network Analyst does not have a solver to do this as efficently as you may want to do it. An attempt can be made by using the Vehicle Routing Problem solver with a set of constraints to do it but may not be what you want. So I would not recommend you getting Network Analyst for it (but you could probably an evaluation copy to try out if needed).

The best product out there for this kind of work is by a company called RouteSmart and it works in conjunction with ArcGIS. You can read more about that here: http://www.routesmart.com/publicworks.aspx

Regards,
Jay Sandhu
Reply
0 Kudos
Highlighted
New Contributor
Our research here is directed toward the snow plow routing problem.  We have developed some artificial intelligence based solutions that are proving to be effective.  Such non-deterministic methods allow solutions that take into account minimization of u-turns, prioritization of arterial roads, route allocation, etc.  We have published a couple theory papers and a trade journal report.  We also are starting to work with local highway superintendents.  At this time we are actively seeking other collaborators!!!  Please let me know if you'd like to discuss this more.

Jim Zollweg, jzollweg@brockport.edu
(Associate Professor of GIS, Earth Sciences)
Reply
0 Kudos