What Does Problem-Solving Look Like? Part 2 of 2

Dear Gemba Coach:

Thank you for explaining the different types of problem-solving that distinguishes lean. But how can I tell that we really are learning the best way to deal with problems? What are the signs of progress when it comes to problem solving?

Good question - and thank you for your willingness to embrace your problems! Beyond distinguishing various kinds of problems, is there anything specific to lean in the way problems are addressed? I believe there is. Here is a quick list:

1. Look for leadership: to a large extent, lean systems are on-the-job training systems. You can tell that you are tackling problems properly when people are taking ownership of their problems. There should be daily training of operators about standards through "stop the line and confirm every problem;" and second, there should be constant training of management about problem solving. Consequently, for every problem the first question will be: who needs to learn to solve this kind of problem right now? And then: whom can we develop by giving them this problem to solve?

2. Visualize the problem: this is unique to the lean approach. Before attempting to resolve the problem, the lean technique is to make the process visible (in real, material terms on the gemba) so that the problem will be revealed when it appears. Properly done, the visualization should highlight the gap between the standard and the actual conditions.

3. Try things out right away: experimenting is not about solving the problem quickly but about understanding it better through trying out ideas quickly. Just as with visualization, there is a unique lean skill in knowing how to try ideas safely and without cost (for a short period of time, with mock-ups, etc.). The key to trying things out is to do it with frontline employees because they know more about the real situation than any manager ever will.

4. Involve all stakeholders: this does not mean that every one's wishes will be granted. But it does mean that everyone involved will have his or her opinion heard. If we can't deliver on what they'd like, we need at the very least to explain why and how.

5. Check rigorously before you draw conclusions: a sign of true lean problem solving is the creation of an indicator that measures how well the subject of the test is affected by quick experiments. Checking is planned in the early steps of PDCA and carried all the way through to the final conclusions.

6. Generalize by transferring the analysis method rather than the solution: in this, lean practice differs considerably from the usual spreading of "best practice." Because the lean approach requires a more detailed understanding of problems and problem solving it is quite commonly acknowledged that a solution that works in one set of conditions might not work in another (certainly adopting Toyota's solutions hasn't done many companies much good). The trick is to spread the analysis method. Experience shows that if people with a similar problem attack it following problem-solving steps which have worked elsewhere, they will come up with their own variation of a generic solution - and make it work profitably and sustainably.

When I first started studying lean initiatives 15 years ago, the veterans used to tell me that problems were great opportunities for improvement. I remember thinking cynically "come on, who in his right mind would believe such blatant company propaganda?" But I’ve come to share the same sentiment. Problems are improvement opportunities because they are occasions to learn about our own processes. Lean is a system to empower people by teaching them how to solve their own problems on the job and, in doing so, develop sounder judgment and better decision-making. The issue is not to "solve" problems every day by making them disappear, but to become more proficient at solving more difficult and more detailed problems every day.

Post a Comment (1)

What Problems Are Lean Practices? Part 1 of 2

Dear Gemba Coach,

I keep reading that lean is about solving problems. But that's exactly what I already do in my job. So how is lean different?

That's a very good question. To explore this issue more deeply, let's start by clarifying exactly what we mean by "problems." Let's distinguish four specific types of problem solving: (1) the workaround, (2) the immediate countermeasure, (3) structured problem solving and root cause analysis, and (4) the kaizen initiative. The distinction I am making here is somewhat arbitrary as there are other problem-solving situations, as well as overlap between these categories, but let's use it now, for argument's sake, to try and clarify the question.

The workaround is what we tend to do intuitively when something goes wrong. We don't want to solve the problem; we just want it to go away. So we look for the simplest way we can go back to our "normal' work. We harvest a missing part from one machine to compensate for a missing part in stock, for example, which does not solve the problem but does postpone it. The overhauled equipment will be immobilized until the replacement part arrives, but this part may arrive before any damage is done, so no one is wiser. Workaround works as the "garbage can" model of decision-making. Problems are represented as garbage cans that land on your desk. You chuck into them the obvious solutions that come to your mind, and this eventually makes the problem disappear (the can vanishes). If you can't find the right "solution,"you can then toss the garbage can on your neighbor's desk, who will then try the same thing. In the lean perspective this is not problem solving. So: what is problem solving in the lean context?

Identifying Gaps

Lean defines a problem as a gap to a standard - a discrepancy between the way things should be running and the way that they are. To have problems, you need standards. Otherwise you have concerns or issues, but not problems. Standards are clear descriptions of good working conditions, which enable gaps to be immediately identified and thus addressed. These types of responses are immediate countermeasures. For instance, the warehouse holding the parts will have clear visual controls that clearly indicate that the components that should be there are actually there. If a slot is empty, workers can call for the part before it is needed for an urgent repair. This is what lean calls an immediate countermeasure. The lean manager does not wait for problems to explode in her face because she is constantly worrying about whether her operations are in "normal conditions." This means work that follows standard processes, which orients her to looking for gaps and reacting immediately when seeing that something is not right. Safety problems highlight this very clearly. It is easier to monitor whether someone is wearing a seatbelt, or not drinking before getting behind the wheel, than it is to try and take the steering wheel out of their hands to avoid a fatal crash. Yet, this requires (1) a precise understanding of what it means to drive safely and (2) the personal ability (self confidence, political clout, etc.) to face the person when they tell you that they don't need to wear the belt just for a couple of miles, that they only had a few glasses and will be fine, or that keeping to the speed limit makes them drowsy.

Immediate countermeasures should be standardized as well. As you deal with problems you will encounter common issues that appear frequently and which can be addressed with common responses. For instance, a customer quality complaint can trigger the standard response of immediately installing a 100% check at the end of the process. Remember that the goal of immediate countermeasures is to (1) protect the customer and (2) return to normal conditions. The great difficulty is to resist the workaround under pressure. Clearly something needs to be done in a hurry, but it is also important to follow procedures. Only high-risk and unique circumstances should trigger a "red alert" awareness with a quick workaround. Yet workarounds should never be considered normal or acceptable: they are the occasion for "all hands on deck and battle stations," and employees should be especially on the lookout for unexpected consequences.

Root Cause Analysis

The next set of situations warrants a deeper kind of problem solving. These are the situations that can't be fixed easily, which cannot return to normal conditions quickly. One example is a work condition that frequently deviates from the standard. The problem repeats itself often, and while workers know a countermeasure, actually using it disrupts work. This would include using special freight to get parts from the supplier when they're not in stock. We know how to use these urgent transport means, which let us live through the problem but they are costly and disruptive. And they are not the proper action, which is to investigate the causes of recurring out-of-stock parts.

Such situations warrant a deeper kind of problem solving: one that strives to identify the root cause of the problem in order to enact the right solution and avoid the usual blundering of trying one thing after another until we stumble on a satisfying compromise. Problem solving of this nature is above all a test of leadership. So when you are confronted with this process ask yourself who has the leadership ability to conduct the analysis to the end and follow the method smartly as well as engaging the other stakeholders.

This structured approach to problem solving has been clearly explained by John Shook in Managing to Learn and by Art Smalley and Durward Sobek in Understanding A3 Thinking. I urge people to read both these terrific books. Essentially, it comes down to following the PDCA approach through eight steps:

1. Clarify the problem

2. Grasp the situation

3. Set a target

4. Analyze the root cause

5. Develop countermeasures

6. See countermeasures through

7. Evaluate both results and process

8. Standardize successful processes

Finally, there is kaizen. Kaizen is about improving processes even when there are no fires to put out. Because no process can ever be considered perfect, there are always opportunities to learn and improve through careful analysis and experimentation. Kaizen exercises train the process owner to understand his or her operations better by tackling a particular problem. There are several standardized kaizen methods that include 5S, one-piece-flow, TPM, SMED, value-stream mapping, and more. These are standard analysis method one can use on any process to understand it better and unearth opportunities for improvement. One can then quickly try solutions until the performance and the process is improved. Kaizen is first and foremost a tool to teach managers to run their processes more effectively and to listen to the idea of their employees.

I hope you’ll see that these last three approaches - immediate countermeasure to return to standard, root cause problem solving, and kaizen - are significantly different from a simple workaround. The rigor and discipline of these problem-solving approaches are essential to lean. Lean leaders must master the process of problem solving. Recently, a CEO of a midsize company who is driving a lean transformation through his business told me that he was facing a challenge. He was arguing with his top management team. The CEO was struggling to explain to his top managers that he expected his middle management team to improve their processes, not to simply keep them running on a daily basis. Making processes run was the job of frontline management, not mid-level managers. And mentoring middle managers so that they understood this was the job of senior managers. In order to improve processes, the leaders needed to teach middle managers how to solve problems rather than invent workarounds for their subordinates. In actual practice this is a radical change of the professional culture. Ask yourself: is this really what you do when you solve problems at your company?

Post a Comment (1)

Do Managers "Get" Lean?

Dear Gemba Coach,

You mentioned during the webinar that plant tours reveal how well managers 'get' lean. Can you be more specific?

Certainly. One of the key purposes of practicing "go and see" is to examine the current state of gemba leadership. After every workplace visit I will ask senior managers how well they believe the site manager understands lean. What do they "get"? What don't they? How can we help them? So during the workplace visit, we look at three things:

1) How lean are their processes (how much muri, mura and muda are visible)?

2) How far are they along in terms of visual management and kaizen efforts?

3) How do their actions reflect their understanding of the lean principles?

The best measure of their lean understanding can be seen in the physical aspects of their operations - regardless of whether they use the tools or not, or even whether they display a keen intellectual grasp of the system. For some lucky few, this lean stuff is just common sense. Like great athletes they can just do it without necessarily being able to explain why they are performing so well. A second measure is how hard they work at lean. Implementing (and maintaining) good visual management requires work, and so does keeping up with a steady stream of kaizen activities. Thirdly, do they do it right? Can you see a clear link between the amount of activity and improvements in processes and performance?

Now, over the years I've learned that the most important aspect for managers to understanding is . . . people involvement. Succeeding at lean involves two fundamental, interrelated areas: continuous improvement and people involvement. Without continuous improvement, people check their brain outside the workplace and dumbly follow processes someone else has thought up. Without people involvement, none of the improvement possibilities demonstrated by the lean approach can hold over time. Lean processes won't stay lean if all people at work are not deeply engaged in constant improvement. In places that haven’t been leaned out yet, it’s easy to improve productivity and reduce costs quickly. But working with leaner processes demands more thoughtfulness from the people who run them. So if one progresses on improvement without involvement, just like skipping on one leg, sooner or later you fall flat on your face. Conversely, involving people without giving them a way to express their engagement will raise expectations without actually improving experience, and disappoint all in the end. These two aspects of lean are inseparable. You can do one without the other, but not for long, and not without painting yourself into a corner.

Most plant managers who do lean understand the value of seeking improvement as a way of fighting the natural entropy of operations and the unavoidable decay of any equipment. Managers may be disappointed by kaizen that doesn't immediately generate large gains, but, as they practice, they recognize that small streams make large rivers, and that a constant flow of small step improvement is necessary to avoid falling back to old ways. What they don't get, however, is involvement. What makes me say that? Let's look at a few typical shop floor cases.

Kaizen events: an obvious stepping stone into lean is organizing frequent kaizen events. The first sign that the plant manager is committed to lean is that he or she takes the time to come to the final presentation where the group explains the improvements made. There's usually a discussion about the improvements themselves, and how significant they are (can they be costed right) and so on. But some questions are glaringly missing from the discussion.

1) First is muri (unreasonableness): are operators happier working in the cell after than before the kaizen? Have ergonomics or simple annoyance been improved? Have we asked the people themselves? Usually the kaizen group is comprised mostly of technicians with a few token operators who can discuss the why and wherewithal of improvement actions. So no one talks to the team actually performing the work. It’s very rare to find a plant manager who goes straight to the working team and asks: is the work easier to do now than before? Which of your suggestions have been implemented?

2. Second, mura (stop-and-go): kaizen groups mostly focus on eliminating muda, which is great. But after the event, go watch the cell for half an hour and count the number of times that operators are interrupted in their normal cycle. Changes are made to make people more productive, but this is offset by the fact that they are still pulled out of their cycle every three parts. This is like having to write a difficult report under deadline pressure, and the phone doesn’t stop ringing. After making the work environment safer, we need to focus on making it more pleasant by taking away interruptions.

3. Thirdly, muda (waste) elimination: sure, at the end of the event, many things have been changed that should make the work more effective. But have we done what is needed to make it stay this way? Have we established the management system to let people prevent things from slipping back? Have we taken the time to let them understand the changes made? The acid test of any kaizen event is returning to the process two months later. This is hard to take, but too often, absolutely nothing is left. Why? The people and their management haven’t been involved in figuring out what is necessary to keep a higher level of efficiency over time.

Stable teams: after creating cells that deliver a set value stream (a fixed basket of products), you must staff them with stable teams: the same operators working together in the same cell everyday. The benefits in terms of productivity are clear. When we measure a stable team against an unstable one (because of frequent temps, for instance) we typically find a 15% productivity difference. So why can't managers achieve this? Because it's hard. Customer demand is often hard to level, for one, so management reserves the right to move any worker anywhere in the plant in the morning according to immediate needs. Once, while looking at the pph numbers with a plant manager, we noticed that one of his three shifts had significantly higher output. The team leader had a ready explanation for this: she refused to take new operators in her team. Interesting, instead of drawing the right conclusion about teamwork and stability (as in 18% higher productivity), the plant manager later complained to us that that woman was totally inflexible and a real pain. This was not a bad plant manager. Indeed, he'd done some spectacular kaizen workshops and was implementing pull. But he simply did not "get" the people part: people involvement delivers productivity and quality.

Operator training: we all agree that frontline operators actually make the product or deliver the service, so they add the most value. Yet how much operator training do we actually carry out? I frequently ask operators: when were you trained at this station last? Too often, the answer is: "when I joined." Most managers see training as an obligatory cost without any real return. Bt contrast, many Toyota tools are easy to misinterpret if you don't see them for what they are: training tools. I have to confess that for many years I saw andon (pulling the chord to light a board which calls the team leader) as a tool to increase management productivity. I was in a Toyota plant once, and told my tour guide so. He looked at me, puzzled, and said "no: operator training." I didn't get it, so I repeated: management reactivity. "No," he insisted. "Operator training." So in the end, I had to ask about operator training. What he explained is that Toyota works hard at persuading all team members to pull the chord whenever they have a doubt. The team leader then comes and discusses their interpretation of the standard both in terms of quality (is the part good or not) and of standardized work (are they following the correct work sequence or not). So every andon pull is a training opportunity. If the team leader doesn’t pull the chord again in the space of a couple of minutes, the line then stops and this becomes a tool for management reactivity. But stopping the line is not that frequent, whereas andon pulls happen every minute or so. Involvement again: in this case I didn’t get it!

Quality circles: these days, every one laughs at quality circles. "Oh yeah, that thing from Japan we tried twenty years ago? Never worked." The only people I know who don't find quality circles funny are Toyota mid-level managers. Quality circles are a large reason why they have so much work. Quality circles start with a stable team investing about two hours overtime a week from some (or all) members studying detailed issues in their area. Why do they do this? Surely not for the great cost benefits. They do this for the involvement. Why don’t we see working quality circles in most places outside Toyota? Because managers don't get people involvement and aren't ready to pay for it.

What, then, is the ROI in involvement? Frontline workers make hundreds of products all day long. To make these products well they've got to follow a repetitive routine and stay focused on doing the work right all day long. This is difficult when they are constantly interrupted, or when they are doing a mechanical task all day long with little emotional commitment. Involving people through kaizen has two huge benefits. First, it helps people pay attention to their work. And second, it engages them to work more smoothly, with less burden and fewer interruptions. Finally, they are encouraged to look for anomalies and spot problems before they become full blown fires.

Most managers who commit to lean eventually do 'get it.' Unfortunately, managers often realize the "continuous improvement" aspect of lean but not the "respect for people" dimension. And so they hobble down the street on one foot, occasionally falling flat on their faces. Lean tools cannot be fully understand without considering both aspects. Hourly parts boards are "production analysis boards" which guarantee hourly production and generate operators comments about what went wrong. Kanban cards insure that the right products arrive at the right place in the right quantity, and give operators a clear waiting queue of products to make so that they can work smoothly and see what's ahead. Standardized work is about getting operators to adhere to a set sequence of steps in a set time, and about hearing all their difficulties on the workstation and spurring suggestions for improvement or "positive variance" (the one way to make the part better than the standard). Without seeing the twin aspects of key lean tools, people can easily get them badly wrong, and then wonder what went awry with their lean implementation. So be careful to look, and to see, these types of signs at the gemba.

Post a Comment (0)