ESD/Mechanical Engineering Seminar

Manufacturing Systems Engineering at Peugeot – Efficient Factory Design and Improvement Using Operations Research

By Alain Patchong,  Project Manager, PSA Peugot Citroen Technical Center Velizy (Paris), France

Abstract:
PSA Peugeot Citroen, the sixth-largest car-maker in the world and the second-largest in Europe, is growing rapidly.  PSA will launch 26 new car models between 2003 and 2006, and production in 2006 is projected to be 4 million cars, up from 3.3 million in 2003.

To meet this objective, PSA decided to focus on its car-body shops, which were the bottlenecks of its plants. An R&D team conducted a project to improve car-body production for PSA. All PSA's cars are now manufactured on lines designed and continually improved with the team's analytic operations research tools. These tools have significantly increased throughput with minimal capital investment, contributing $130 million to the bottom line in 2001 alone.

This presentation describes one of the most recently developed tools, which allocates buffers and robots in car body welding lines. It also shows how analytic tools are helping PSA staff understand and convert to lean principles.
The car body is the naked, unpainted steel shell of the car (also called "body in white" in the US).  It is assembled in shops where most of the operations are performed by robots that load and weld stamped steep parts. These robots are organized in modules (groups of robots working on the same part) and separated by buffers. One of the main objectives of car body shop designers is to keep cost as low as possible. To do that, they have at their disposal two main levers: the numbers of buffers and robots. Adding additional buffers could reduce the impact of disruptions such as failures and, consequently, increase the production rate. On the other hand, adding robots will speed up the lines which would also increase the production rate. Both add significant but different costs. Also, additional robots means additional failures, and this may reduce or reverse the increase in production rate. Given a target production rate, the goal of this method is to help production line designers answer the following questions: What set of robots and buffers will meet the target at least cost and where should they be installed?