» » next post - Michael Ballé: Pick to light and learning to teach Jidoka
« « previous post - Art Smalley: Jidoka – Part 1
Art Smalley

Art Smalley: Jidoka – Part 2

By Art Smalley, - Last updated: lundi, août 9, 2010 - Save & Share - Leave a comment

As I mentioned I’ll share a recent story about Jidoka from a client visit and discuss what we learned in the process of implementation.  Usually when people talk about Jidoka the first examples that are discussed involve an operator on an assembly line pulling a chord which stops the line. Then a supervisor comes running and fixes whatever problem just occurred. I don’t really consider this full blown 100% true Jidoka however by my standards…

Jidoka is when a machine (not just a human pulling a chord) is enabled with some type of device which senses an abnormality which in turn automatically shuts down the process from further operation. Defects and other problems (e.g. evidence of the crime scene) are thus kept in station. The concept is the opposite of “automatically run” and instead is “automatically stop” unless of course there is no abnormality.

Sakichi Toyoda’s looms had creative mechanical devices which caused the machines to stop running when a thread broke as in the following case. Unfortunately without motion it is difficult to appreciate but the following image is of a mechanical sensor from nearly 100 years ago. The breakage of the horizontal thread (weft) causes a metal finger to drop down and stop the machine.

1) Machine running normally

2) Thread breaks. Finger drops and machine stops.

I challenge my clients to move beyond just superficial lean and the Just-in-Time pillar of the production system. On a recent visit I had an opportunity to review some of the Jidoka attempts and what was learned.

One team was struggling with about 5 percentage points per shift of downtime on an automated assembly line. There was a station where five steel bearings were being inserted automatically into an inner race. From time to time a ball was not dropping into the race. That caused problems and the need for rework down later in the line when it was finally detected.

The team realized this was an opportunity for Jidoka and attached a simple sensor with count capability. Unless five balls were detected then the process stopped and someone had to attend to the machine. For a while this did nothing other than prevent the problem from going further down the line. After a couple days however it lead to improved problem solving…

Certain pallets appeared to have more problems with a ball not inserting than the others. It took some digging and 5 Why thinking but finally the team realized that the drive gear that engaged the pallet  from under the machine was not engaging with some pallets and this lead to some initial spinning and balls not hitting the inner race as intended. The teeth on the drive gear and its mating surface were both square teeth by design. We suggested altering the profile so that the teeth on one side would be slightly triangular and more likely to engage right away each time. After several attempts with trial and error a suitable shape was found that worked better and soon no more stoppages of this kind were reported.

Jidoka in this fashion is a a tremendous aid in problem solving and getting to the root cause of an issue. The concept of stopping the defect in station ensures that more evidence is available for problem solving. I don’t know how many teams I see trying to solve problems simply by deducing things out of thin air. Unless you are the intellectual equivalent of Albert Einstein – good luck! The best chance you often have of solving a problem is at the location where it occurs, when it occurs, and by examination (including measurement) of the actual parts in question. In Toyota this is referred to as “GenchiGenbutsu” and in conjunction with Jidoka it is a tremendous aid in getting to the root cause of problems. In this case it wiped out a problem that had been happening for years as far as I know.

I don’t think there is any one magic thing that makes the Toyota Product System work. However if you can put in place hundreds of small Jidoka concepts such as this one you’ll be a much more successful company than if you had not.

Post to Twitter

Share this post...Tweet about this on Twitter
Twitter
Share on LinkedIn
Linkedin
Buffer this page
Buffer
Share on Facebook
Facebook
Email this to someone
email
Pin on Pinterest
Pinterest
Share on Tumblr
Tumblr
Posted in Uncategorized • Tags: , , , , Top Of Page

Write a comment

*