© 2010 IBM Corporation IBM Container Transport Optimization Dr. Juergen Koehl, IBM Research and Development, Boeblingen Juergen Koehl, IBM Container Logistics 19 March 2010

© 2010 IBM Corporation Introduction  Container transport volumes are increasing considerably over time  Many operations planning models do not scale with this increase  Efficiency is gained with sophisticated decision support systems  The optimization potential of multi-modal transportation costs comes –from the possibility to choose between different transport modes –combine import and export of containers to minimize the empty tour segments  In the case of empty container re-positioning additional saving come from the different to/from location options as well as the accuracy increase of the forecast

© 2010 IBM Corporation Challenges  What is the cost optimal combination of transport modes and carriers?  What is the trade off between cost and service level?  How much equipment do I need? And where? And when?  How do I react to disruptions of the operations? How do build a robust plan?  Can I accept this transport request?

© 2010 IBM Corporation Empty Container Re-positioning  The model: –Input: Network: Node and Links Scheduled and on demand connections Demand: Per container type –Decision variables are the choice of Mode, carrier, time, location, number and type of containers –Constraints: Transport capacities Container demand for each location Storage capacities Business rules –Cost function to be minimized: Transport costs Storage costs Equipment costs Nodes Links

© 2010 IBM Corporation Empty Container Re-positioning  The model: –Input: Network: Node and Links Scheduled and on demand connections Demand: Per container type –Decision variables are the choice of Mode, carrier, time, location, number and type of containers –Constraints: Transport capacities Container demand for each location Storage capacities Business rules –Cost function to be minimized: Transport costs Storage costs Equipment costs Nodes Links Linear Model ∑ size(c) ≤ cap Minimize Transport cost + storage cost + handling cost

© 2010 IBM Corporation The solution: IBM ILOG Optimization Decision Manager (ODM)  ODM is the industry leading platform for developing custom optimization-based planning and scheduling applications.  CPLEX is the leading optimization engine for linear optimization problems

© 2010 IBM Corporation Integration into the Customer IT 7 ODM Optimization Server ODM Repository ODM Studio (Planner) Integrated, Customized Web GUI Existing Application Data communication GUI uses component OPL Studio Developer

© 2010 IBM Corporation Benefits of the Solution  Highly automated planning process  Exploit low-cost operations alternatives and counter- intuitive low-cost decisions  Supports strategic analysis: –Optimize the number of containers needed –Reduces the size of the container fleet  Reduced annual logistics expenses of empty container repositioning by 5 – 15%  Implemented in ILOG industry standard optimization technology –Provides an optimization platform for future solutions  Supports incremental planning to react to disruption  Fully customized to the existing IT Infrastructure, business process and requirements

© 2010 IBM Corporation Other Applications of Container Transport Optimization  Overseas container transport and re-positioning  Re-positioning of empty rail cargo freight cars: ~ 100 different types of freight cars ~ 1000 ‘nodes’ or freight train depots Scheduled cargo train connections  Container transport in the automotive industry  Forecasting of container demand