Dispatching Shipments

Dispatching Shipments

Once a production and distribution network has been designed and put into operation, daily operating decisions need to be made about how to send the shipments. Some of these decisions again are yes-or-no decisions.

For example, suppose that trucks are being used to transport the shipments and each truck typically makes deliveries to several customers during each trip. It then becomes necessary to select a route (sequence of customers) for each truck, so each candidate for a route leads to the following yes-or-no decision:

Should a certain route be selected for one of the trucks?

The objective would be to select the routes that would minimize the total cost of making all the deliveries.

Various complications also can be considered. For example, if different truck sizes are available, each candidate for selection would include both a certain route and a cer- tain truck size. Similarly, if timing is an issue, a time period for the departure also can be specified as part of the yes-or-no decision. With both factors, each yes-or-no decision would have the form shown next.

Should all the following be selected simultaneously for a delivery run:

1. A certain route,

2. A certain size of truck, and

3. A certain time period for the departure?

For example, one BIP application of this type was developed by Petrobras, the largest corporation in Brazil and one of the world’s oil giants. As described in Selected Refer- ence A7, Petrobras transports approximately 1,900 employees daily between about 80 off- shore oil platforms and four mainland bases, using more than 40 helicopters. A BIP model requires less than an hour to generate optimized helicopter routes and schedules each day, resulting in annual savings of more than $20 million. Thus, the “shipments” being dis- patched in this case are groups of employees.