A very Lean global automotive Original Equipment Manufacturer had to very quickly increase production volumes because increased demand had created a backlog

Context

This case study is about a mission Marris Consulting did in 2011 in Mexico for a large automotive equipment manufacturer, which was reputed to be among the 30 best practices in terms of Lean Manufacturing in the world. The factory was about 1,000 people broken up into 20 autonomous production units.

Our client’s need

Our client asked us to help them in one of their factories in Mexico. Their problem was that they were not managing to produce enough. There was a fairly low demand for their products in previous years and suddenly the demand boosted up. They started having a lot of trouble to deliver their clients on time and to honor their orders. They asked Marris Consulting to try and help them by injecting Theory of Constraints into their system to see if we could produce more in that way.

Objectives

The goals were:

• To increase very significantly the production volumes to meet a big increase in demand

• To eliminate the client delivery backlog

• To put an end to huge Premium Freight expenses (airplane deliveries)

• To restore the client confidence with its clients especially their biggest customer who is one of the largest car manufacturers in the world

Approach used / Process

First, we went to the factory and we did a rapid unambiguous identification of which part of the 20 production units was bottlenecks. We identified two and a half of them as being bottlenecks.

To build the product, a dozen of operations is needed. The bottleneck was one of these operations, in a production unit which was about 10 people and produces about 6,000 units a day on a 24-hour basis.

They had already formally identified their bottleneck correctly with a sign next to it saying "this is the bottleneck".

So what we did is implement one of the rules of the Theory of Constraints about 1 or 2% of the whole principles of the Theory of Constraints and that improved the throughput of the performance of the unit by over 15% in 15 minutes.

How did we do that? What rule did we use?

We just added a buffer here in front of the bottleneck.

And in 15 minutes what we did would just change the rules. We said they should always be at least four parts there, that is to say four times 15 seconds about a minutes worth of work. And that was enough to absorb all the micro stoppages that were going on this unit. That ended up in fact being 17% increase within they are on a stable basis simply because they were getting lots of stoppages from this resource which was a non-bottleneck but not that reliable and therefore we were losing quite a lot of time there.

The approach in video: How to increase productivity by 15% in 15 minutes ?