110 likes | 229 Vues
This lecture covers essential concepts in routing and scheduling for transportation logistics, focusing on the Traveling Salesman Problem (TSP) and Vehicle Routing Problem (VRP). Key considerations include determining which customers each vehicle should visit, the order of visits, and cost minimization while complying with constraints such as truck size and driver hours. Real-world examples from online grocers like Peapod and Webvan illustrate the practical application of these theories in managing deliveries, route optimization, and improving logistics efficiencies.
E N D
Transportation LogisticsCEE 498B/599I Professor Goodchild 4/18/07
Routing and Scheduling • Which customers to be visited by each vehicle? • Sequence in which they will be visited • When will each customer be visited? • How much will this cost? • Minimize cost and travel time, missed deliveries, comply with constraints (truck size, driver hours)
TSP and VRP • Traveling Salesman Problem (TSP): Given N points, find a tour that visits them all, returning to the point of departure, with minimum distance (here we are not dealing with many vehicles) • Vehicle Routing Problem (VRP): Find an allocation of points to vehicles, and a set of tours that return to a depot, that minimize either distance, number of vehicles, or a combination of the two (this is the same problem we addressed with the SMM and GA method, but number of vehicles not given)
Text Example: Peapod • Online grocer • Each day has different realization of deliveries • Webvan • Example: • distance on grid proportional to actual distance traveled • Cost proportional to distance traveled • Capacity 200 units • Minimize distance traveled
Location/Allocation • Segment the area (assign customers to a route) • Decide on route
Sequence Customers on Routes • Determine an initial route (not necessarily optimal) • Farthest insert (add farthest from DC first) • Nearest insert (add closest to DC first) • Nearest neighbor (add closest to customer just added first) • Sweep
Route Improvement Procedures • Can we do better than original tours? • There are many ways you could do this: • 2OPT: break trip 2 spots, reconnect • 3OPT: break trip 3 spots, reconnect
Transportation Networks • Direct Shipping • Milk-run • Distribution center (hub) • Tailored network (serving different customers with different modes, methods or frequencies)
Trade-offs • Management complexity • Inbound/outbound transportation costs • Cost of facility • Inventory • Handling • Ability to respond to the customer • Mode
In-class exercise • What attributes of a logistics system make milk-runs appropriate? (e.g. high density customers, large loads) • What attributes make a cross-docking facility appropriate? • When does it make sense to aggregate inventory?