Group of Numerical Optimization and Modeling. GNOM

Strategical Models in Supply Chain Design

Strategical Models in Supply Chain Design through Stochastic Programming

Partners and members

  • Group on Numerical Optimization and Modeling, UPC (Barcelona).
    • Dr. F.-Javier Heredia, professor, project leader.
    • Daniel Ramón Lumbierres, PhD. student.
  • Accenture TechLabs (Silicon Valley).
    • Mary Hamilton, Managing Director.
    • Sunny Web, Senior Principal.
  • Accenture Analytics Innovation Center (Barcelona)
    • Robert Gimeno, Managing Director.
    • Elsa M. Sans, Digital Strategy Consultant.

Introduction and objectives

The Optimal Supply Chain Strategy (OSCS) is a stochastic optimization problem for help manufacturing companies deciding (1) which operations and processes would compose a supply chain, (2) which speculation-postponement strategy would fit better and (3) the optimal production and stock levels for that strategy, in order to maximize the total benefit of the system and subject to capacities, time and risk constraints.

The uncertainty of this problem is introduced by a scenario tree with multiple stages. At each stage, demand levels for its time periods are assumed known, while remains stochastic for future stages. All stages are composed by a set of time periods, where production decisions are taken. Against this backdrop, the speculation-postponement strategy refers to decide the flow of production

  • spatially i.e. to decide the forecast-driven production (speculation) and customer-driven production (postponement); and
  • temporally i.e. to decide whether to move certain quantities of production forward the chain or to hold them until future periods.

Therefore, a critical decision is to decide which active operations will allow the storage of finished or semi-finished production, setting the breakpoint separating speculation and postponement strategies. Each one of these activities are known as Customer Order Decoupling Points or simply Decoupling Points.

This research project is the follow-up of the one-year project “Digitalizing Supply Chain Strategy with 3D Printing”, led by professor F.-Javier Heredia (GNOM-UPC) and professor Joaquim Minguella (FundacióCIM-UPC). Accenture Labs, the Accenture’s center for advanced developments in Silicon Valley, and Accenture Analytics Innovation Center of Barcelona have settled this project with GNOM-UPC to continue the research of this topic, the main objective being the optimal design of supply chains that includes the speculation-postponement paradigm though multistage stochastic programming. 

Meetings

2017
Top

Production

2017
2018
Top

 

Conferences

2018
  • 23rd International Symposium on Mathematical Programming, July 1-6, Bordeaux, France.
  • 29th European Conference on Operational Research, July 8-11, València, Spain.
Top


Papers

 2017

 2020


Keywords
supply chain stochastic programming