Fleet Routing Optimization

Someone have worked in a Fleet routing optimization problem in Knime?

I am in a FMCG company, that deliver items to small shops around the city.
We have the coordinates of our clients, average times per visit, and a fleet of vehicules to satistify our customers, so i want to optimize the routes in order to reduce costs.

Hi @AR7 -

I moved this topic to the main AP forum to give it a bit more visibility. :slight_smile:

I’ve worked in routing optimization, and personally I wouldn’t use KNIME to try and accomplish the goal. Not a knock on KNIME, but in my experience, the amount of variables involved to truly optimize a route can be overwhelming.
For example…
While it’s relatively easy to set a route to minimize distances, there are WAY more variables to consider, such as…

  • -Is the business open or closed? (e.g. special holidays)
  • -What are the delivery time windows for each business for each day?
  • -Can the business handle trucks with lift gates or no lift gates?
  • -How soon do you want to revisit the same general area?
    • E.g. Say you create a route on 2 consecutive days. Each day, independently, has an optimized route to minimize the distance between each stop. However, on the second day, you deliver to a customer directly across the street from a customer you delivered to on day 1. While it is true that each day had an optimized route, if you look over the course of the 2 day span, it’s very likely that it was unoptimized, as you went to the same area 2 days in a row.

If you JUST wanted to minimize distances between stops, you could probably set up a Traveling Sales Person (TSP) problem with ease. However, as mentioned in my last point, it would give you optimized routes for a single day, but to REALLY create efficiency, you need a forecasting component, and you need the TSP to look at delivery need over several days.

Again, not a knock on KNIME, but specialized software is probably worth the investment. I would love to be proved wrong however :slight_smile:

2 Likes

If you just want to try out the TSP problem @Snowy mentioned you could take a look at this example and discussion:

3 Likes

Thank you for your answer @Snowy .

I think you understand very well our business problem.
Every day can have a different optimized route, because 95% of the clients are visited once a week. And it’s necessary to consider all the variables that you’ve mentioned and to consider the average sale per client to have profitable routes.

What would you recommend me to solve this business problem?

Best regards!

Thank you @ScottF !!!

@AR7 I’ve used Roadnet by Omnitracs in the past. Solved many of the challenges outlined and is fairly flexible when new challenges arise. Good luck to you!

2 Likes

This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.