The VRP Solver can be configured to solve many different fleet routing problems. What are you using it for?
I would like to solve, given:-
Determine the optimal (start) locations to place the field force.
I'm working with Vehicle Routing Problem (VRP) on a fleet of vehicles configuring, Orders (Time Windows, Service Time, Routes (Delivery Quantities, MaxTotalTime, MaxDistance) Route Seed Points (Dynamic) to cluster a lot of orders, but this is very dificult because it depends of a lot of variables that you can't controlled, I'm applyng this in a retail organization.
Do you have any idea what I should take in mind? or if I can combine with other analysis that can help me to cluster orders.
For other options on clustering you might take a look at location-allocation and tools in the Spatial Statistics Tools\Mapping Clusters toolbox. This might help break the problem down into smaller groupings that can get better results when doing routing.
I work in public transit and will use the VRP to assist in developing and planning some specialized transit solutions around work and school.
That is a very interesting problem. We don't have too many of our users telling us about using transit with VRP. We'd love to hear more of your workflows and how it is going. In case you do not have this link already one of my colleagues has put together an awesome story map on workflows and resources for transit. http://arcg.is/2timm2E
I was finally able to devote some time to this task. I get to do GIS maybe 5% of my work time in a month. I do have Premium StreetMap Custom Roads, but for this exercise I used a Michigan Roads layer that was built by the state. Rachel Applebaum informed me how I had my ONEWAY traffic patterns coded incorrectly. I am delighted to have this network up and running. I will use the Custom Roads with Pro next, as we will feed turn by turn directions thru Navigator to bus drivers. On to the problem.
Flint and Genesee County in Michigan has few economic opportunities. Surrounding counties have employment, but residents in those communities commute elsewhere for higher paying jobs. The Flint MTA has regional routes that anyone can use (disclaimer: this is not a charter service) and we travel to employment districts in those surrounding counties. Multiple businesses and shift times, both starting and ending, affect the solution. Shift change usually takes 20 min from unloading workers and waiting for the shift to end and those workers coming to load the bus.
I initially thought about using the two different time windows, but I didn't see that I could specify that the 1st start and end time window is tied to the Delivery Quantity and the 2nd start and end time window is tied to the PickUp Quantity.
Every Delivery and Pickup received its own Order. Using the StartTimeWindow1 and EndTimeWindow1 as parameters for when the companies would like employees to be dropped off and picked up, the VRP would not have our vehicle waiting at each location for the shift change.
I completed my first attempt at this solution. I have to tinker with the MaxTotalTime and MaxTotalTravelTime parameters to ensure we are not inconveniencing our riders by having them on the bus for 2-3 hours for an hour long trip.
All and all, my boss is really impressed with this program. This company did not use Arc or any GIS prior to hiring me. I'm slowly introducing the functions and capabilities of the ESRI products as I learn them and can convey how they will benefit different parts of the organization.
I need to know the options that the software offers to increase the occupation of the capacity of vehicles (routes), I have read the VRP Solve bibliography, but I couldn´t find a variable that controls this parameter.
The options that I already worked are:
None of the above gave a positive result.
The software respects time windows, optimizes resources but underutilizes load capacity. An example of this, a vehicle with a capacity of 10 tons dispatches it with deliveries equivalent to 3 tons, a vehicle with a capacity of 3.5 tons dispatches it with 2 tons, etc., for the company that distributes mass consumption products. This is not efficient since the cost of transport per ton rises, even though it respects time windows and other restrictions.
Example: The table on right side represents the results of the vrp, while the table on left side represents the expected results. The goal is to increase the average of %UseCapacity, in this test the costs were zero for all routes.
I would appreciate to your suggestions or study material where there are options to solve this requirement.
For the different routes that you have in the problem are there other constraints that are different between them? Starting/Ending Depot, max total time/travel time/distance, max order count, etc?
Also do you have breaks set up for the routes?
I work in distribution company and will use the VRP to help in developing and planning the route for each vehicle.
We have big and small vehicles.
Big vehicles are not allowed to enter city downtown.
I am reading the documentation as developer, but cannot find where I can define this restriction to big vehicles.
I know I can define route zones where I set limits where each vehicle can go, but I need the opposite, where it can not go.
That sounds like an interesting problem. You could try making the route zone for an area that surrounds the city's downtown and goes far enough out into the suburbs to cover all of the orders leaving a hole in the route zone for the downtown area.
Thanks Heather for your response.
I can do that, I will try it.
I hope that someday there will be a specific solution to this problem,
because big tracks can go far from the city, too.
Now I should define zones as big as the whole country except downtowns
At present, we have two trucks that collect kitchen waste and they need to do the job at night. Each truck has a maximum capacity for 12 waste containers and there are 20 kitchen waste points which have from 40 to 20 waste containers respectively. The starting point is the same as the ending point. When arcgis desktop is used for drawings, two problems happen. 1 the work assigned for certain kitchen waste collection points are from 23: 00 a.m. to 25: 00 a.m., but arcgis does not recognize the timeline as according to it, the latest end point should be 23:59:59, but we do have to work at night.2, If the route is well arranged and a truck is arranged to collect 40 containers of kitchen waste, how can it be arranged to continue collecting the remaining kitchen waste? (For example, when the truck takes away the first 12 containers, how should the remaining 28 be dealt with?). At present, some points cannot be well scheduled because the trucks’ carrying capacity is set up to be 12 containers. If I make their capacity to be 100 containers, all routes will be formed immediately. I have seen some of your topics: 1, for a point with 40 containers, it is supposed to divide them at the same place to orders of 12 containers by 3, plus an order of 4 containers. But when the capacity of the truck is fixed to be 12 containers, it still will not work. I tried to add vechiles in a virtual manner, namely, I added A A1, A2, A3 vehicles virtually to the plan when we only have A and B in reality, and I set the starting point the same as the original starting point, it turned out to be irrational. Do you have any other good way to do this?3 If I am going to set the end point of the recycling process not as the same starting point, but another point in the schedule before 9:30 PM, and then, after 9:30 PM , I want to set the end point back as the very starting point, how should I do it?
This sounds like an interesting problem. I think we should be able to address your three concerns.
1.) When working over night, it is important to add date information as well as time. So for example you have collection points from 23:00 to 25:00 am, if you add date information into that you would instead input this as 01 June 2020 23:00 am to 02 June 2020 01:00. This lets the solver know that it is a two hour time frame that just happens to span across midnight. If the actual date isn't of importance to you just pick dates that are representative of when you normally run these routes such as spanning from Monday to Tuesday.
2.) When a vehicle can pickup a load from orders, empty the vehicle, and then go back out to more orders, we use route renewals to model the emptying the vehicle. For this you will need to add in the location that the vehicles can be emptied into the depot feature class if it is not already in there. Then add in information in the Route Renewals table. You will need a row for each of your vehicles. You can find more information about Route Renewals in our doc, here. It is important to note that the solver assumes that it can empty the vehicle at the end of the day at the ending depot so if you need to empty the vehicles else where, make that location the ending depot.
3.) I am not sure I full understand this question. If you are saying you would like a way to route the vehicles back to the starting location after emptying out the vehicles, that would best be done as a separate solve using the Route solver. If you are asking for a time dependent ending location (point A if before 9:30 and point B if after 9:30) that is not something our solver currently supports. I think the way I would handle this is to leave the ending depot blank (you have to have a starting depot to do this as we require at least one of the start depot and end depot). This will stop the route at the last order location. You could then determine which ending depot you would like the vehicle to return to and determine that path using the route solver. Alternatively you could solve the VRP problem twice, once using point A as the ending depot and once using point B as the ending depot. Then when looking at the results look at which is correct based on your time restrictions.
Hope this helps,
Retrieving data ...