E-Logistics Successes and Failures Copyright April 2002 H. Donald Ratliff H. Donald Ratliff 1 Executive Director Regents & UPS Professor don.ratliff@isye.gatech.edu President & CEO don.ratliff@velant.com
INFORMS The Successes and Failures of E-Logistics H. Donald Ratliff Regents & UPS Professor of Industrial & Systems Engineering; Executive Director of the Logistics Institute Georgia Institute of Technology President and CEO, Velant, Inc. The Internet is revolutionizing the process and economics of logistics. Twenty years ago, the personal computer launched a new generation of logistics technology based primarily on a decentralized planning model. It empowered sophisticated users to make better decisions by giving them flexible modeling capability, a graphical interface and a moderate amount of computing power. Client-server computers added an element of connectivity that allowed automation of internal company transactions but did little to change decision technology or interactions between companies. E-logistics--the next generation of logistics technology--will be based on a centralized computing model driven by networked computing. In this session, Ratliff will discuss the evolution of this new model, the successes and failures to date, and the technical and operational issues that must be overcome. 2
Internet-Centric Computing A network of networks Public & private Wired & wireless Service Center Plant DC Vendor Colocation Center Easy access Computer horsepower Location neutral Carrier Store 3
E-Logistics Technology Matrix Mainframe PC Internet- Client- Server Centric 1960s 1980s 1990s 2000s Monitor Status Internal static Internal dynamic Shared dynamic Execute Transactions Accounting Internally integrated B to B B to C C to C Planning& Analysis Distributed Interactive Hybrid Centralized Collaborative 4 Transportation Planning
Transportation Planning $500,000,000,000 annual US truck transportation spend (3,000,000+ trucks) 2/3 Dedicated Fleets 1/3 Commercial Carriers Fleet Optimization loads, routes, schedules Mode Opt Fleet Opt Shipper/Carrier Transaction Automation 5 In the US there is as much money spent each year on truck transportation than on the combination of warehousing, inventory, and all other forms of transportation
Transportation Planning Network Design Facility location & sizing Channel definition Strategic Procurement Optimization Capacity planning Carrier bid optimization Shipment consolidation Load planning & scheduling Mode selection 6
Transportation Planning Network Design Facility location & sizing Channel definition Strategic Procurement Optimization Capacity planning Carrier bid optimization Shipment consolidation Load planning & scheduling Mode selection 7
Supply Chain Design Software Copyright April 2002 H. Donald Ratliff 8 One of the biggest Operations Research successes
Network Optimization Models Models High level Product flows Time buckets Linear cost Transport cost Plant capacity Plant cost DC capacity DC cost Customer demand Algorithms LP/IP Pretty fast Effective with super-users Transport cost
SC Design Software Current Offerings CAPS Logistics I2 Insight Logic Tools Solver engines ILOG - CPLEX Dash Xpress MP Custom Future Requires high modeling expertise Not much change 10
Transportation Network Design Pooling networks Pool points UPS facility Postal facility LTL hub Cross-docking networks Focus on LTL shipments Cross-docks are shipper controlled Origins Destinations Cross-docks 11
Cross-Docks 12 No planned inventory Unload, sort, stage, load Examples Inbound for retail Wal-Mart Home Depot LTL and package delivery UPS Yellow Freight New cars Ford Operation Schedule-driven Load-driven receiving shipping sort
Design Difficulties Origin to destination TL LTL Multi-stop route Origin to cross-dock TL LTL Multi-stop route Cross-dock to destination TL LTL Multi-stop route Routing is difficult Static Dynamic Supply chain design models don t work!!! Cost are very nonlinear Flows are dependent Volume is critical Origins Destinations 13 Technology Inadequate!!! Cross-docks
Transportation Planning Network Design Facility location & sizing Channel definition Strategic Procurement Optimization Capacity planning Carrier bid optimization Shipment consolidation Load planning & scheduling Mode selection 14
Bid Optimization Red all or nothing Software CAPS Logistics I2 Managed service Logistics.com Bid opportunities 15
Transportation Planning Network Design Facility location & sizing Channel definition Strategic Procurement Optimization Capacity planning Carrier bid optimization Shipment consolidation Load planning & scheduling Mode selection 16
Transportation Optimization Loading A C D Routing & scheduling B A A? B B Backhauls 17
Mode Optimization Versus Fleet Copyright April 2002 H. Donald Ratliff Optimization Mode optimization Create routes from LTL shipments Tender routes only if less cost than LTL Lane based costs Involves an external transaction Fleet optimization All shipments must be taken Cost based on dedicated truck Driver Equipment No external transaction 18 Commercial Carrier Dedicated Fleet
Current Technology Map-on-the-wall Map-on-the-computer 19 Visual algorithms Requires smart planner Poor quality Greedy algorithms Requires super-user Marginal quality
Why is transportation optimization Copyright April 2002 H. Donald Ratliff hard? Huge number of possibilities Complex constraints Loading Delivery windows DOT Compatibility Dynamic requirements Large amounts of data Changing service levels Min E {Σ cx} Ax = b x > 0 20
How can optimization complexity Copyright April 2002 H. Donald Ratliff be overcome? Executive Analysis & Simulations Client Systems Orders Optimized Transportation Plan Transportation Management Center Status Loads Routes Schedules Status Field Operations Order/Asset Visibility 21
Transportation Management Copyright April 2002 H. Donald Ratliff Center Design and Analysis Manager (Off-line Simulation) Orders Customers Assets Roads Status History Data Goals Constraints Rules Business Models Step improvement Continuous improvement Optimization Engines Exception Handling Loads Routes Schedules Results Support Automation Workflow Manager (On-line Execution) 22 Optimization engines are the key to success!
Optimization Providers Fleet Planning Software Service Mode Optimization Software Service Software Hybrid Service Descartes X UPS-Roadnet X CAPS X X X Manugistics X X I2 X G-Log X Logistics.com X Velant X X X 23
Collaboration Levels of collaboration Share data Status Share plans Future availability Future requirements Jointly plan Share assets Share capacity Joint planning technology Interactive planning More people More time Questionable value Composite planning Combine independent plans Automated collaboration Difficult to create Difficult to maintain 24
Network Effects Capacity sharing Visibility Collaboration Optimization Solution providers Elogex Logistics.com Nistevo Velant Limited success Private Fleet - Backhaul Commercial Carrier Continuous Move
Workload Imbalance Copyright April 2002 H. Donald Ratliff 12 10 ATL DAL KC NY CHI DET LA SF Trucks Required 8 6 4 2 What to do with excess capacity? 0 Day
Transportation Exchanges Vision Sell unused capacity Workload imbalance Backhauls Partial loads Bring together shippers and carriers Many small carriers Many small shippers Reduce cost Brokers fees 13% Increase competition - auctions Leverage information Shippers Exchange Carriers 27
Transportation Exchanges More than 100 startups Most are gone Spot market small Too many players Too little experience Weak technology Survivors New business models 28
What about the future? Network design Better cross-dock planning technology Problems are really hard Strategic procurement Mostly automation Trends Centralized optimization Optimization outsourcing Limited collaboration Optimization is too hard for carriers Optimization Transportation management centers New generation of technology 29
Questions Comments 30