0

I have a list of zip codes from customers that need to be visited by a salesrep within a specific timeframe. I can translate the zipcodes to a matrix of distances between each zipcode if required.

Assuming every customer needs to be visited within the same fixed time span.

How can I cluster/assign customers to salesreps so that every customer is visited in within the timeframe and the total number of salesreps is minimised?

Can this be restated as a 'standard' problem?

ErwinM
  • 5,051
  • 2
  • 23
  • 35
  • You will need to define a few rules. Like how much time does a sales rep take with each customer? Is it the same amount of time for each visit? – Marquis Blount Oct 30 '15 at 10:52
  • Yes, its the same time per visit. In a way i am trying to divide an area with dots (customers on a map) in the least amount of circles. Where the circle diameter is fixed (determined by traveling distance and visit time). – ErwinM Oct 30 '15 at 11:03
  • I think the problem as stated is heavily under-specified. Could you give a sample? What is your exact input? What do you expect as ouput? What rules does a salesman have to adhere to? are there routes? is this a geometric problem or are you rather looking for a TSP-thing with multiple salesman? Its not clear to me way you couldn't simply say: saleman needs `x` min per person. There are `n` person. The timeframe is `t`. `n*x` is the total time which will be needed to process the `n` persons. Thus you will need `n*x/t` salesmen. Obviously its not what you mean: what's missing? – dingalapadum Oct 30 '15 at 11:55
  • Your calculation would ignore the driving distance between the customers. In essence I am trying to model what would be an optimal way to visit every customer with the least amount of sales rep given a geographical distribution and fixed time frame. – ErwinM Oct 30 '15 at 13:49

0 Answers0