What? My Pull System Is Supposed to Fail?
Dear Gemba Coach,
We’ve given up on installing a pull system – every time we try, our on time delivery rate plummets, and we can’t let customer service suffer more than it already does. We’re doing a lot of A3 problem solving – isn’t that lean enough?
A pull system is supposed to crash – that’s precisely what it does. A pull system is not a device to replace the MRP with cardboard cards. It’s not a production system technique. It’s a device to improve the production system, and to learn. The tension between the need to produce exactly at takt and the commitment never to pass on wrong or unfinished work is the workplace for kaizen. The confusion is easy to make and the vocabulary doesn’t help. Lean is about:
- Improving customer satisfaction in terms of quality, lead-time and cost: this can’t be done by simply carrying work as usual more rigorously but actually examining every failure to deliver quality on time to figure out the errors in thinking that caused the mistake in the first place.
- Tightening the just-in-time conditions of production by reducing lead-times from customer order to delivery, from production order to shipping, from supplier order to production, by further leveling, fractioning and mixing production in order to get closer to takt time – this typically involves making production flexible to reduce batches and improve flow towards continuous one-piece-flow.
- Spotting defects closer and closer to where the defect is produced and stopping production rather than accepting a defective (or incomplete) part, making a defective part or passing on a defective (or incomplete) part or order.
- Involving team members more deeply in improving their work areas and mastering standardized work by working at a smooth repetitive flow and taking away all the reasons for unnecessary movements and efforts (waste).
Typically, when Toyota starts production of a new model, very few cars come out of the production line at first because they are stopped along the process as issues are identified. Very soon, most of the problems are solved and the challenge is to achieve target cycle time as soon as possible. In normal conditions a line should run somewhere between 94% to 98% of target. When the line gets to 98%, resources are taken away, which leans the process further, reveals new problems, and the line falls back to 94% and so on. This constant tension is what keeps the leaning process alive and kicking.
What kind of issues a pull system is likely to uncover? Typically, we’ll find 4M problems:
- Material supply issues: missing parts are an endemic difficulty in any complex assembly, particularly with many options. The pull system withdraws from supplier processes the parts needed just-in-time and, eventually, makes the supply flow smoother particularly as assembly is scheduled for leveled pull on components;
- Machine availability issues: certain processes prove troublesome particularly when they’re over complex or inflexible. Machine tools designed to perform too many operations on the part, presses too hard to change and so on.
- Manpower issues: in many cases, a lack of stable teams means operators are moved around in the shop and not trained well enough at handling certain processes they can be unfamiliar with – this creates spot problems for them and for the flow that are hard to pinpoint as long as people are not stabilized in the shop and don’t develop ownership for their own processes.
- Method issues: if you stand just 10 minutes in any assembly area, you’ll see that although the overall flow of work is often defined as soon as you look at detailed assembly operations the team members fight against ambiguity and hurdles. Parts are jumbled together in boxes and need to be unraveled, screws are poorly positioned and difficult to put in, components don’t fit well together and need hammering in and so on. These many problems cause quality issues and late deliveries, as well as endless team member safety or ergonomics problems.
Now, of course, falling from 98% to 94% is hardly “plummeting” and if attempts at setting up the pull system really have dire consequences, such as an OTD lower than 70% as I’ve sometimes seen, then, of course, you should stand back and rethink your efforts. In early attempts at pulling, it is easy to fall in any of the following traps:
- Not leveling well enough – pull can hardly happen all at once from scratch – it has to be learned progressively. The trick is to focus on the very high runners, pull the first, then the second and progressively go down the list by 1) first leveling the demand of the every high runners over the week (the law of high numbers makes it likely that the percentage of daily demand variation on high volume won’t be too bad) and 2) pull progressively the products with higher peaks and deeper troughs. The weekly levelled planning meeting is essential in learning to smoothen a variable customer demand.
- Sluggish logistics pull – pull means, well, pulling, which is picking up products or parts throughout the day – at the very minimum once an hour. People are often tempted with one or two pull pitches a day to make it easier on logistics. In doing so they forego the “pull” effect on the line, don’t improve much and discredit the whole pulling idea which now looks like another corporate raindance.
- Being hostage to a process – this can be a “bottleneck” troublesome machine, or one that needs setters to change (and the setters are never available) or a supplier that simply can’t ship the needed components. Pull can reveal one key part of the flow that is hard to improve because it’s owned by someone who won’t or can’t improve. This is a particularly difficult case as it often creates hard political battles around the idea of pulling.
Chasing Your Tail
A3 problem solving is of course beneficial but the real question is: what is a problem? How are problems picked? What kind of solution is expected? Human beings naturally favor problems they know how to solve whereas the very function of pull is to make you face the problems you have to solve – even if you have no clue how. The risk you run by pushing A3 problem solving without pull is to run around in circles chasing your own tail and never addressing the deep issues people are working around. As Chris Argyris once pointed out, many organizational dysfunctions stem from taboo issues, and the fact that the issue is taboo becomes taboo as well. Pull breaks through all of this fancy footwork.
Pull is never easy because it always challenges the status quo. On the other hand, pull progressively makes sure all the rowers in the boat row at takt, in rhythm with each other, which is what makes the process perform.
The root of many problems with pull is to see it as a production technique. It’s a management technique. Pull is the main tool of the site director or even the operations manager to get the flow to behave the way they want it to. It’s a dressage technique for a very powerful and willful horse to take the image of The Lean Manager. Mistaking it for a production technique which will advantageously replace the MRP is an error few recover from because it completely underestimates the political pushback a real pull system will create as long as you continuously increase the level of just-in-time and take cards out. Pull is an essential tool if you who take the “pursuit of perfection” seriously. I don’t believe “lean” makes much sense without pull.
How can there be standards -- or kaizen -- in a service job when no two instances are the same?
Dear Gemba Coach,
I’m in a service job and I struggle with the idea of standards. I read that there can be no kaizen without standards but how can you have standards when no two instances are the same?
I don’t get kanban -- I don’t work in production so how would it apply to one-off work?
Dear Gemba Coach,
I feel that I still don’t get kanban. I don’t work in production, and I fail to see how stock replenishment would apply to one-off work.
Can lean be used to turn around a company?
Dear Gemba Coach,
Can lean be used in a turnaround situation? Does a burning platform make it easier or harder?