Latest Contributions by HeatherMoe

For order pairs, the first order must happen before the second order. This means that in you example you should never see the Depot1(Pickup)1 before its corresponding Order3(Pickup) since they are paired together with the order being the first of the pair. What you could see is Depot1(Delivery)1 --> Order3(Pickup) --> Depot1(Pickup)1 --> Order1(Delivery).    A parameter that might be useful for you to adjust is called Transit Time Importance (found in the Advance section in ArcGIS Pro) or referred to as excess_transit_factor in the REST API. This will influence the allowable additional transit time between the first order and the second order. If you prefer the route go directly from the first order in the pair to the second then set this value to High, or if you want to allow extra time by picking up additional orders before going to the second order in the pair then set this value to Low. When set to Low the main focus is on creating an efficient route so the amount of time between the first and second order in the pair is less important. Medium tries to strike that balance between amount of time on the vehicle and overall efficiency of the route.    Some other thoughts on how this could be handled depending on the size and complexity of the problem: 1. Using the specialties for corresponding to the starting depot and the ending depot with the orders having the matching designations. Then create duplicate routes at each starting depot for each of the possible ending depots. When given extra routes then are needed the solver should only pick-out the best needed routes for the solution.   2. Solve the delivery problem and the pickup problem in two different solves. For delivering to orders from a required start depot the routes can have a virtual ending depot. Similarly for picking up at the orders and delivering to a required end depot the routes can have a virtual start depot. Then once both problems are solved you can pair together the routes and then resolve to get an optimized sequencing with preserving the route. ... View more

With both the travel time and distance slightly shorter, this seems like it is just a product of the solver being a heuristic and by passing in different parameters it is going through a slightly different set of steps.  ... View more

The VRP solver uses a heuristic meaning it does not find an exact (best) solution every time but instead follows a series of steps to get a good answer. Those series of steps change based on the parameters that are set in the problem and can result in the solver finding a different good (local optimal) solution. By changing the cost of the distance it affects the total overall cost the solver is trying to minimize and so it might (and in your case does) find a different local optimal solution then when the solve is run without the distance cost because it is solving a different problem.   If you look at the output from those four solves you might also see that the total distance traveled is shorter for patterns 3 and 4 and that the shorter distance cost offsets the slightly longer travel time. ... View more

@dylan_onfleet For balancing on distance you might try a similar approach of solving the problem first to understand what the cumulative total distance is for the problem. Make sure to use the CostPerUnitDistance field on the Routes since the distance is an important parameter for your use case. Once you know what the total distance used by all routes you can determine the average and set the MaxTotalDistance value on the Routes to something slightly larger then the average based on the acceptable variability between routes. You will want to make sure to give a MaxTotalDistance value larger then the average so that there is enough flexibility in the routes to continue to route everything. It is also worth taking a few minutes to look at the unbalanced distance solution from the first solve to see what other constraints the routes are running into when they are particularly short or extra long distances. It could be that some of the short routes are spending a lot of time at orders working and so evening up the number of orders by lowering the MaxOrderCount could help.  ... View more

Hello Rasmus, We have a few videos that might be helpful for understanding how to set up your problem. Both videos work through the same demos but this one uses ArcGIS Pro and this one uses ArcMap depending on what you have available.  From your description above my question would be how much of the straw bales is at each of the storage facilities? If it is above the level that can be carried by a single vehicle, duplicate orders will need to be used at that storage facility with the quantity broken down into groupings that are capable to be picked up by a single vehicle. For example (making the numbers easy) if a truck can carry 10 tons and there are 37 tons at the storage facility. Then four orders would be needed to represent that storage facility. Three of them would have a pickup quantity of 10 and the fourth a quantity of 7.  Hope this helps, Heather ... View more

This sounds like an interesting use case and not something I've heard before. I don't think this is something fully supported by the VRP solver but something to try would be using two sets of order pairs, one going from location 1 to location 2 and one going from location 2 back to location 1. If on the routes you also add a small ArriveDepartDelay it will encourage the solver to stop at orders at the same location consecutively and by the same route. The ArriveDepartDelay will add time to the route for each stop the vehicle makes. If multiple orders are at the same exact location and visited consecutively the time is only added once because the vehicle only arrives and departs once for all of those coincident orders. This should encourage the drop off of the first order pair and the pick up of the second order pair to happen consecutively. It is certainly not a guarantee but maybe it will get most of the round trips correct and only a few would need adjusting by hand.  ... View more

Can you check the status messages for that solve? I believe it should be giving an error message about the orders in the order pairs being reused. An order can only be in one order pair so using the same orders for the order pair that connects the delivery to the pickup should be what is causing the problem. ... View more

I'm trying to understand your problem a little more. It seems like you are wanting one driver to pick up someone/something at location 1, drive to location 2 and then back to location 1 with that same person or item. Is this correct?   Is there a certain amount of time at location 2 that the person or item needs to stay there? and if so can the driver go to a different location to pick up a different order during that time or does the full round trip need to happen before the driver can do anything else?   If the driver is allowed to do something else while waiting for picking back up from location 2 is it possible to then have two orders on the vehicle at once or would a full pickup and drop off need to be done during that time?   When expanding the problem to multiple routes, is it a hard requirement that the return trip is done by the same vehicle/driver or could that be done by a different route in the area?     ... View more

As mentioned above the Orders and Routes have ViolatedConstraints fields that should provide some more detail about why it is not solving. The Status field on the orders might also give a clue if they are not locating correctly. ... View more

Thanks for the suggestion. We are in the early stages of discussing how to handle datetime fields for network analyst layers and will consider this requirement as we move forward with those talks. ... View more

The method you describe of making multiple copies of the routes for each day is what is recommended for this type of problem. The solver will only use the number of routes that are needed and so by adding more then is expected it should help you determine how many routes are needed.  ... View more

Hi Nelson, I have been able to reproduce the selection not being honored by the solve VRP GP tool. We are looking into fixing this. In the mean time if you are working within the app my best suggestion is to save a temporary copy of your feature class with just the portion you are interested in for the solve. And if you are working within a python script the Network Analyst module (arcpy.nax) will honor the selections when solved. I, however, am not able to reproduce the output pickup quantities showing as zero instead of showing the number you have for inputs. If you continue to have this issue, please open a support ticket.  ... View more

Hi Ihab, From that warning message it looks like you also have orders with an assignment rule of preserve route or preserve route and relative sequence. True? If so then no we do not have a mechanism for continuing the solve with the remaining orders. When orders are given these assignment rules we first check to see if the route is feasible and if not error out because something is likely wrong with the problem set up that was not intentional from the user.  ... View more

With adding layer and ribbon support to VRP in ArcGIS Pro 2.6 we also updated the schema of the solver. More details on that can be found here. Although the Solve Vehicle Routing Problem Tool did not have any changes to the tool functionality, it now goes through a bit of translation code to change the schema to the new format to send to the solver. We have seen some changes in the outputs between nulls and zeros as a result of this. If you are seeing functional differences please report these so we can look into them. And if you are working within the ArcGIS Pro application, I would encourage you to look into using the layer and ribbon capabilities we added for VRP in ArcGIS Pro 2.6.  ... View more