Halt and Catch Fire
Our series on Data Analytics and Change Continues. In this week's post we travel back in time to discuss Lean, Six Sigma, Aliens, and the Ford Pinto.
Read time ~7 min.
By Jakob Klaus, MBA, MSW, and Principal Partner for Tenacious Change.
The year 1986 was not a great year. It started with the Challenger Space Shuttle disaster. By summer, a large portion of eastern Europe was contaminated with fallout from the Chernobyl disaster. By the Fall we all learned why the Weyland-Yutani Corp had lost contact with the colony on LV-426.
Okay, that last one is the plot of Aliens (1986).
Still, it was a rough year, yet for those interested in systems thinking and process improvement, it turned out to be quite a promising year.
In November of ‘86, Japanese business consultant Masaaki Imai published his first book in the Kaizen trilogy, Kaizen: The Key to Japan’s Competitive Success, which introduced people to the concept of Kaizen, better known today as Lean.
At the same time, Bill Smith at Motorola, Inc was inventing the Six Sigma method for quality improvement. Lean and Six Sigma are complementary processes and today are practically synonymous.
Currently Lean and Six Sigma are used by many manufacturing and service companies to improve quality and reduce costs. Remember that EVA (economic value added) can be increased through increasing revenue and/or by increasing efficiency to reduce costs. Lean & Six Sigma do just that by using efficient production and service processes with a focus on constantly improving quality and efficiency.
Lean was a response to the traditional “push system” for production, in which a product is pushed from one workstation on the production line to the next without concern for whether the next workstation could use all the parts produced at the first workstation.
For example, let’s say it’s 1971 and Ford expects high demand for their newest subcompact car, the Ford Pinto. The management team takes the forecasted demand for Pintos and translates it into a production schedule.
Under a push system the upstream workstations (the start of the production line) will start producing as many widgets as required for the Pinto regardless of whether the downstream workstations (those closer to the end of production line and the customer) can use them that day, week, or month.
This results in a “work in progress” (WIP) inventory and the lack of connection between one workstation and the next results in increased costs for storing WIP parts.
Meanwhile, at Toyota the management team has forecast the demand for their own compact car, the Toyota Corona. Toyota sets a production schedule but is using Kaizen/Lean system which pulls outputs from upstream suppliers to downstream workstations.
With Lean, the workstations closest to the customer set the pace of production and pull new parts from upstream stations only when needed. This produces cost savings by not having a WIP inventory, but also has another benefit. A pull system has a low tolerance for poor quality parts. The parts you get are the parts you use so if you get a bad widget that’s a problem that has ripple effects across the entire production system.
This is where Lean, Six Sigma, and systems thinking come together. The Lean system requires greater connectivity across all assembly processes. So, when a widget is too big or too small to use at the next workstation, the entire system can come to a halt until the problem is resolved. Six Sigma reduces variation in the quality of parts to ensure the quality parts are available when they are needed.
This system also works well in the service industry. Just ask Ronald McDonald.
One of the fascinating aspects of Lean & Six Sigma is that it can tell you a lot about process behavior and the nature of systems change.
In Gemba Kaizen a Commonsense Approach to Continuous Improvement Strategy, Masaaki Imai describes the quality improvement process at Yokogawa Hewlett-Packard (YHP) from 1977 to 1982. This is illustrated in a great graph on page 42 of Gemba Kaizen which I do NOT have permission to reproduce. ☹
Starting in 1978 YHP had a failure rate of 4,000 parts per million (ppm), by 1982 the failure rate was down to 3 parts per million. That’s remarkable progress, but the takeaway from this is that progress frequently did not look like progress.
More than half the time the system’s behavior got temporarily worse after a process improvement. YHP implemented 9 process improvements to reduce the failure rate. Following 5 of those improvements the failure rate temporarily increased.
The first phase of the quality improvement involved implementing the Lean concepts to improve the workspace, train the team, and introduce new equipment. During the first phase the failure rate dropped from 4,000 ppm to 40 ppm. The failure rate only increased once, and it was a doozy. After introducing new equipment, the error rate doubled, erased the progress of the past 4 months, and did not return to its prior ppm level for another 7 months!
I imagine the Quality Engineers at YHP had a Private Hudson moment when they saw that.
The second phase involved YHP refining the process to get below 40 ppm. During this phase, every improvement resulted in a temporary increase in the failure rate, but the failure rate continued to toward fewer parts per million.
The progress at YHP illustrates what a lot of us who use systems thinking in our day jobs have observed: sometimes regression is a prerequisite for performance.
Got a business data question for Jakob? Send him an email at jakob@tenaciouschange.us.
Now Playing on Getting to Third Space with Lamar & Tom
Watch the podcast now on the Tenacious Change YouTube Channel or listen on Spotify.
Be sure to subscribe to Change It Up! and be among the first to know when new episodes are available.
To learn more about our work with communities, organizations, and groups, visit us at our company website at www.tenaciouschange.us by clicking on the image below: