» » next post - Art Smalley: Jidoka – Part 1
« « previous post - Jean Cunningham: Poka Yoke in an administrative process
Michael Balle

Michael Ballé: Jidoka is the key to on-the-job learning

By Michael Balle, co-author of The Gold Mine and The Lean Manager - Last updated: jeudi, août 5, 2010 - Save & Share - Leave a comment

I remember visiting Toyota’s French plant and standing in front of andon board: the call lights kept flashing on and off. An operator would call the team leader, who would sort the problem out within the imparted time before the fixed-point system would stop the line. “Management reactivity,” I said. “Nope,” they answered, “operator training.” “This isn’t used to get management to react faster to problems?” I insisted? “Operator training they repeated.” And so on.

It took me a while to understand I was projecting our usual management models on Toyota’s practice. In my worldview, management’s role was to be there to solve every problem – the “inverted pyramid”- which was already a step forward from ignoring problems and working on financial powerpoints. But still, my idea was that the system should increase the management of the line.

The Toyota engineers had a different perspective on it. To them, the only way to deliver the required level of quality at the required takt time to satisfy customer expectations was to make sure that operators got EVERY cycle right. Yes, the line stop will stop the line if the andon call issue can’t be solved by the team leader – and yes, a countermeasure will then be found by management, and hopefully this will trigger a deeper “5 why?” analysis. But the main point here is how do we avoid having to stop the line. In other words, how do we increase the capability of the line.

There are two aspects to improving quality: capability and containment. Capability is the propensity to get parts right first time. Containment is the ability to stop bad parts from moving to the next step in the process. Jidoka’s lesson, to me, is that the key to improving capability is containment. Operators (or, for that matter, engineers on machine intensive processes) will be able to know more about what they do if they can call whenever they have a doubt, and be taught the difference between OK and not-OK. This on-the-job, ongoing training is essential to building superb capability. It’s one by one confirmation of the work done.

I have to confess I have rarely seen actual andon outside of Toyota. But the first step of Jidoka is creating a simple space for a conversation between the team leader and the operator on whether the parts are OK or Not-OK. So we start with a red bin. There is a dedicated (red) spot for operators to place any doubtful part, call the team leader, and together decide whether the part is OK or Not-OK and start to backtrack where things went wrong. The system only works, of course, if frontline management is interested and keeps checking what comes out of the red bins – the next step might be a “marketplace” where bad parts are organized by type of defect and so on.

The key to improvement is being able to distinguish a good part by a bad part – whether directly on the machine through simple “poka-yoke” like systems, or by the operator eyeballing the part. What we usually find out is how far we are from being able to do this. Red bins fill up so fast it’s almost impossible to respond to every call, operators find it hard to spot the difference between a good part and a bad part because there are no visual references, the systems engineers invent to make machine defect-proof are so awkward to handle they tend to be disconnected by the shopfloor staff in order to be able to produce, etc. Worse of all, many managers consider that a certain amount of defects are unavoidable and just a fact of life.

How much is Jidoka worth? Well, many industrial companies I come across (this is less true in first tier automotive suppliers) struggle along with 1 to 2 percent of their sales sunk in costs related to non-quality (without even quantifying the commercial costs of sending defectives to customers). People consider this is a very small proportion of costs when,  say, purchased parts and material make 70 percent of costs, or labor twenty. But on the other hand, in many operations, this is equivalent to the margin – and is totally avoidable! Pure muda.

Just by making sure management, as a team, reviewed what was in the red bins once a shift and followed the countermeasures action plan one plant reduced its scrapped parts from 1,4% of sales down to 0,3% in six months – in a low margin industry, a result that showed on the bottom line. And this was just the start.

Without quality, there is no customer trust (I’ve been on a sales call where the salesman was trying to increase price because of materials costs increases, and was greeted by angry shouts from the purchaser because of defective products: an uphill battle!). Without customer trust, no new volumes. Many sites feel that their situation is far too difficult immediately to worry about such things. They want cost improvements now. Few realize that fixing quality is the right place to start to solve both the immediate cost problem and the mid-term competitiveness issue

Post to Twitter

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

Write a comment