Standardized Work in Machine-Intensive Processes
Dear Gemba Coach,
Most lean literature and case studies to date focus on assembly type manufacturing which utilizes very people-intensive operations. This is not the case in the machine-intensive process industries and therefore has major implications on the format of standardized work. Can you shed some light on what standardized work should look like in the process industries?
Thank you for this question, which provides an opportunity to investigate the point of a lean tool. Many people use magical thinking when adopting tools—assuming that by simply applying them, the process will improve and so will performance. Five S is typically prone to such hopes—and disappointments. Hopeful improvers see endless cycles of 5S drives fail to generate any radical change in either process or performance, because the tool has not been applied for the right reasons.
The crucial lean principle to keep in mind is that lean tools were invented to help people analyze their own work patterns and see the muda they themselves generate in their own processes, so that they can look for better ways to get the job done. “Standardized work” (a set sequence of steps to do within a target cycle time), for example, is a great tool to help assembly operators see that they rarely perform the same tasks the same way, see that this involves a lot of walking, fumbling and creates defectives as well as safety risks, and learn a better way. Consequently, you are right to question whether this applies to machine shops or flow processes. The fundamental question is: which aspect of the work do we want to focus on?
The Core of Standardized work
Before we get to the nitty-gritty of the tools, let’s go back to the Gemba. I recently visited an industrial tannery where raw hides are processed on a conveyor through chemical baths, and then dried in large oven-type equipment. Or, a few years back, I worked with a manufacturer of industrial carpets: huge rolls of materials go through a number of processes in large equipment that fills up entire halls. Where are the operators in such settings? Some operators work at set stations in the process, typically to move materials from one step to the next. For instance, the hides will be taken off the conveyor to be taken to the ovens. This could be automated, but in some cases the flexibility necessary for the operation requires human labor. Operators can also be found doing specific operations such as tool changeovers or material supply. In a machine shop, for instance, an operator’s main job is to bring the part to the machine, insert it and then set the proper tool for machining. This would be true also of the large flow-processing operations such as the carpet making I mentioned. Finally, human operators are also found running the entire process – most commonly within a central control room where the computer system keeps track of all the equipment, and from there, human operators go and perform a number of checks on the machines themselves.
The core idea of standardized work is to determine the most efficient (as in: “muda-free”) work sequence and to repeat it exactly in the same way so that operators avoid unnecessary motion and wasted effort. Standardized work guarantees quality and precision, saves time, maintains safety and prevents equipment damage. But what about when there is no sequence, because a few operators are called to and fro to deal with materials and equipment, but without any apparent cycle to their work?
Standardized work is primarily there to show the operator a muda-free way of working: by referring to the operations standard sheet and the standardized work sheet, operators can find out what they do wrong and correct it, or what they do better, and improve the standard. Secondly, standardized work is essential in helping operators cope with change: change in the mix of products with tool change-over, change in the volume, with more or less people in the shop (and integrating part time colleagues, for instance), change of production with new product introduction and so on. Without standardized work, each of these changes becomes a huge headache that the shop will naturally try to avoid or resist. Standardized work is an essential aspect of flexibility.
A few operators are isolated at workstations doing either checking or moving components from one process to the next. With all the focus on the machinery, these workstations have rarely been looked at in detail, and are in dire need of being improved.
The periodic jobs which need doing on the machines such as changeovers, basic maintenance, or material supply are themselves muda-prone, as its felt that every time is different and there is no standard way to perform the job. Certainly, any SMED workshop in a machine shop will reveal exactly how unorganized and difficult most tool changeovers really are.
The control part of the process is also rife with muda, as operators move to and fro in large halls to check this or that according to what appears on the computers screen, often after the fact and when they deal with a lot of defectives or rework. Having operators spend their time “babysitting” the equipment is an obvious form of muda in most process industries I’ve seen.
Last – but not least – defectives are often considered as “normal” in process industries: every time tools are changed in a machine shop, or material is changed in a process shop, some parts and some product is considered defective for “setting” purposes. In machine shops, typically, the first part of a batch will be thrown away as part of the setting process. If one reduces the batches to only a few parts at a time to follow real demand, this procedure becomes unbearably wasteful.
#1 Form of muda
The number one form of muda in a machine shop occurs, not surprisingly, when machines go wrong: either they produce defects, or they break down. To avoid this, labor is utilized to “patrol” the shop and make sure the product is good and that machines run. Most of these activities are either too early (you check but there’s nothing wrong) or too late (something went wrong, and you only find out when you come to check). In effect, the way to fight this source of waste is to be able to spot real-time where you need to be when. The main tool for improving machine-intensive processes is andon, before standardized work.
How can the machines spot something is going wrong and how can they call your attention to it? Jidoka is often seen as the hidden part of the lean iceberg – an engineering technology Toyota has developed to make its machines “autonomous” in the sense of having some judgment over whether they were making parts to specification or working correctly. Think, for instance, of installing captors behind cutting tools to make sure the tolerance are held and to stop the machine if they’re not. andon (the call mechanism) is essential to waste reduction in a machine shop or a flow process because it takes away the “random walk” of operators checking on the machines. The equipment no longer needs a babysitter – it will call for help when needed. In this manner one can dramatically reduce the resources needed to ensure things run, and invest more in maintenance activities and problem solving to keep the equipment at its nominal value.
The second common form of waste found in machine intensive environments occurs when operators carry out technical tasks such as changing tools or maintenance. Standing in the Ohno circle during these activities it becomes painfully obvious that much of the time is spend going back and forward looking for tools, parts, second-guessing, correcting previous work and so on. Operations standards (the bedrock behind standardized work) are an essential tool to help people minimize muda in these operations. By (1) listing clearly the steps of, says, the maintenance, (2) highlighting the critical point at each step and (3) explaining why this is a critical point, one has a tool both to help operators learn how to do this better, but also to kaizen their own work by analyzing the specific reasons they do not follow the standard at any given time.
Thirdly, a lot of waste occurs as operators shuttle materials from and to the machines. These trips are mostly organized “on demand” and often suffer from conflicting priorities and many mishaps. On this kind of waste, standardized work (a stable sequence of steps) can be very to help operators better organize their delivery and pick-up routes.
I have discussed this question with Art Smalley (author of Creating Level Pull and co-author of Understanding A3 Thinking), who is currently writing a book on standardized work with Mr. Isao Kato the original creator of the modern day standardized work course in Toyota, and his take from his experience in Toyota’s machine shop is that:
“The question is sort of tough to answer and the devil is in the details ... the 100% strict truth is that Toyota does not use pure standardized work in engine or transmission machining lines. The reason is that the work pattern is not cyclical. Instead the operator walk pattern to different stations is dictated by the andon signal. It tells them to go and change cutting tools, conduct quality checks, and find out why machines have stopped, etc. So in the purest sense it is not governed by the three elements of takt time, work sequence, and standard work in process (SWIP). That is mainly for man machine combination that have repetitive work that is the same cycle after cycle.
“In machining we use operator instruction sheets that for all intents and purposes look the same to the untrained eye. They still contain columns on the left for Main work Elements of the job and a column for Key Points for each step of the job. There will be a picture or diagram to the right and the usual information at the top. If you look closely however the elemental times don't exist and there is no indication for Standard work in process (SWIP). In machining we use lots of operation instruction sheets for tool changes, quality checks, gauging of parts, and various tasks such as machine start up, shut down, or some cases involving troubleshooting.”
To answer your question, the “standardized work” tool was created to help assembly operators find the best muda-free way of assembling parts. In describing the sequence of steps to achieve in a takt, standardized work is a great tool to learn the proper sequence, and then spot problems in the way we currently do things (if we can’t get the job done with the target cycle time, if we don’t do the operations in the correct sequence, etc.); which then highlights the opportunities for kaizen. In a machine shop or process flow environment, the issues are different, and the first question to solve is how to direct the attention of operators in large, complex and demanding areas. Before addressing standardized work questions, make sure you’ve done enough work on the andon mechanisms. Modifying machines so that they recognize the problems they create is never easy and often a worthwhile challenge.
Secondly, when operators do intervene on the equipment – how do we make sure they do it right? Such activities are usually different from walking around a workstation to use equipment in sequence, but certainly a step-by-step approach can help tremendously, as well as a focus on explicit “pay attention” points at each step (which can often differ from machine to machine”. In this case, operations standard would be the more relevant tool.
In specific cases – the change-over is now down to a few minutes, or there is a materials pick up and delivery route – the standardized work sheet in its classic form of showing the operators’ steps from one station to the next can be helpful to find more detailed waste and work on it. But it wouldn’t be where I’d intuitively start. I’d first make sure operators can tell by themselves when to be where, to do what, and how to go about it to do a good job of it. The key to tools is that they should help the people doing the work better understand how they do their job, and see what kinds of muda they are generating in the process. Different situations do call for different tools.