/n Total Points: 7.5

GEOG 683: Hazards and Emergency Management

Lab 2: Vehicle Routing

In this assignment you will be examining how best to route vehicles. Complete the lab utilizing the

12-5 Routing vehicles from depots to demand points data that has been uploaded to ELMS.

Deliverable: insert your completed maps into a Word document and include a paragraph describing

how efficiently routing vehicles from depots/central location to demand points will aid in examining

your disaster (note, if it does not apply to your disaster, use another disaster).

Routing vehicles from depots to demand points

Consumers want to buy locally grown produce (fruits, grains, and vegetables) in their grocery stores,

but grocery stores cannot take deliveries from local farmers directly because of the small volume of

produce a single farmer may erratically have versus the large volumes of produce reliably needed by

the stores. It's too costly and there are too many uncertainties for large grocery stores to take delivery

from many individual farmers. The solution is to provide aggregation depots, which take deliveries

from farmers, large and small, and then deliver truckloads of produce to the grocery stores. That's

just the sort of problem for the Vehicle Routing algorithm of Network Analyst; it minimizes costs for

the depots of delivering the produce to grocery stores. As you will see, the Vehicle Routing

algorithm is rich in terms of complexity handled and parameters you can set for minimizing routing

costs.

Get started

Suppose that the Food Hub Collaborative is looking for a produce aggregation depot in San

Francisco and plans to use up to four dock-high delivery trucks, each with a payload capacity of

20,000 pounds, to supply San Francisco's over 40 major grocery stores with locally grown produce.

For planning purposes, you'll assume that eventually all the grocery stores will be customers with

typical demands based on sales volume. That way you can determine which of the two alternative

depots is best and if four trucks is the right number.

1. Open Tutorial 12-5.mxd from the Maps folder and use the Service Area bookmark.

2. Save the map document.

If necessary, click Customize > Extensions, click Network Analyst on, and click Close. Also click

Customize > Toolbars > Network Analyst.

Add depots and orders

You'll select the Vehicle-Routing model, and then load facility and demand points. You'll select

Sushi Umi and use it as the depot. Later, you'll try the US Trading Company as an alternative site.

1. Select the Sushi Umi depot with the Selection tool on the Tools toolbar.

2. On the Network Analyst toolbar, click Network Analyst and select New Vehicle Routing Problem.

Open the Network Analyst window. 3. In the Network Analyst window, right-click Depots, click Load Locations, and type or make

selections as shown in the image, with SFGroceryWarehouses for Load From, 9:00:00 for

Time WindowStart 1, and 5:00:00 for TimeWindowEnd1.

TimeWindowStart1 and TimeWindowEnd1 are the hours of operation for the depot. For Sushi Umi,

it's 9:00 a.m. to 5:00 p.m.

4. Click OK.

5. Right-click Orders and click Load Locations. Select SFGrocery Stores for the Load From field and

complete the form by making selections as shown in the image, with Address for Name, ServiceTime

for ServiceTIme, TimeStart1 for TimeWindowStart1, TimeEnd1 for TimeWindowEnd1, and 0 for

Max Violation Time 1. Although the field is not shown in the image, scroll down and select Demand

for DeliveryQuantities.

The solver needs unique names for each order point so you'll use Address, which is unique whereas

Name repeats because chain stores have the same name. The ServiceTime field provides data on how

long a stop is (in minutes) at a grocery store to get to the unloading dock, unload, and get back on the

delivery route. Time WindowStart1 and TimeWindowEnd1 provide the time window during which a

grocery store accepts deliveries. Each store has a window of 9:00 a.m. to 5:00 p.m. The default value

of 0 for Max ViolationTimel indicates that time windows cannot be violated. Finally,

DeliveryQuantities is the number of pounds of an order for each store.

6. Click OK.

That adds 43 orders to grocery stores to be on the route for deliveries.

Add routes

Some details are needed for properties of routes and trucks; namely, the capacity and costs of

operating trucks. Fortunately you can create one route (and truck), and then copy and paste it for the

other three trucks because the Food Hub Collaborative would have four new, identical trucks. Then

the Vehicle Routing Algorithm will create the four best routes to meet demand.

1. In the Network Analyst window, right-click Routes and click Add Item.

2. Select or type the properties from the table shown (leaving properties omitted from the table with

the default or Null value).

Value Description

Property

Name Truck_1

StartDepotName

EndDepotName

EarliestStartTime

LatestStartTime

Capacities

CostPerUnitTime

$0.40 per minute

CostPerUnitDistance

MaxTotalTime 390

minutes

3. Click OK.

Name of the truck

Sushi Umi

Sushi Umi

8:00 AM

8:00 AM

The truck starts at this depot

The truck ends at this depot

The truck can start loading at 8:00 AM

The truck must start at 8:00 AM

20000 Capacity of the truck in pounds

0.40 The truck driver is paid $24 per hour, so cost is $24/60 minutes

=

1.25 The cost of fuel, truck depreciation, maintenance, tires, etc. per mile

The work shift for a driver is 8 hours, but total work time is 6.5 hours, or 390 4. Right-click Truck_1 in the Network Analyst window and click Copy. Right-click Routes (1) and

click Paste three times.

5. Edit the three copies of Truck _1 to rename them Truck_2, Truck_3, and Truck_4.

Set properties for the routing problem

As is usual with Network Analyst, next you'll select analysis settings.

1. Click the Properties button in the Network Analyst window > Analysis Settings tab.

2. Type or make selections as shown in the image, with TravelTime (Minutes) for Time Attribute,

Meters (Meters) for Distance Attribute, Monday for Day of Week, Not Allowed for U-Turns at

Junctions, Straight Line for Output Shape Type, and Use Hierarchy unchecked.

3. Click OK.

Solve the routing problem

Now you'll see if four trucks can meet demand.

1. Click the Solve button on the Network Analyst toolbar.

2. Right-click Routes and symbolize with Categories > Unique values using Name for the Value field

and width = 2 for line symbols.

Evidently, four trucks is enough to meet this level of demand, because all 43 grocery stores are

allocated. Next, let's take a look at the level of usage for each truck and costs.

3. In the Network Analyst window under Routes, double-click Truck_1 to open its Properties

window.

Scroll down to see the results of routing. This truck needed only 199.8 minutes of the 390 available

in the workday, only drove about 6 miles for a total of 12.8 minutes of driving, and finished at 11:20

a.m. If you checked the other three trucks, you would see that they used 308, 350, and 371 minutes of

the workday each. Maybe three trucks would be enough to do the job.

Add route renewal

1. In the Network Analyst window, right-click and delete Truck_4.

2. Right-click Route Renewals (0) in the Network Analyst window and click Add item.

3. Make selections or type as follows.

This allows Truck_1 to return to the depot to reload, in an estimated 30 minutes' time, and then go

back out again to finish a route.

4. Click OK. Likewise, add two more items to Route Renewals, including the other two trucks.

5. Click OK and re-solve.

That makes the three trucks nearly feasible with only two stores not getting deliveries. Truck driver 1

works 388 minutes, truck driver 2 works 355 minutes, and truck driver 3 works 382 minutes-getting

close to the 390 minutes available in a regular workday. Not every store would have an order every day, so three trucks might be enough. The tight constraint is MaxTotalTime (390 minutes), so next

try increasing that parameter to 7 hours per day × 60 minutes per hour = 420 minutes.

6. Under Routes in the Network Analyst window, double-click Truck_1 and set MaxTotalTime to

420. Likewise, do the same for the other two trucks and re-solve.

Now all grocery stores can have deliveries.

7. Save your map document.