Unilever’s new and integrated program for World Class Manufacturing
By Dr Ir Jaap van Ede, editor-in-chief business-improvement.eu, 19-10-2015
The three best-known process improvement methods are (1) TPM to increase the stability of the machinery, (2) Lean to increase the value stream towards the customer, and (3) Six Sigma to reduce process variation.
Companies that have been applying process improvement for a long time, often say that they have reached a ‘world class’ level, and speak of World Class Manufacturing (WCM). This means that TPM, Lean and Six Sigma are applied simultaneously, with one of these methods as foundation, and the others added later.
Unilever was a good example of a company with this approach, with WCM based on TPM. However, the add-on approach is a bit unsatisfactory. It is like dealing with shortcomings in your 'house of process improvement', by adding outbuildings.
Therefore, Unilever changed their approach in 2012. Within car manufacturer Fiat they discovered a new and balanced WCM-program, developed by professor emeritus Hajime Yamashima. TPM is the base of this program, as it was for WCM old style within Unilever. However, Yamashima didn't add 'outbuildings' for the other two improvement methods, but he integrated Lean and Six Sigma from the start in the TPM management pillars.
The result: a completely new improvement house, which gives more focus, better opportunities for cross-departmental improvement, and which pays more attention to the role of the people and to the structure of the organization!
Total Productive Maintenance (TPM), the improvement method which Unilever started to apply 20 years ago, focuses on increasing the productivity. By measuring the Overall Equipment Effectiveness (OEE), machine-stoppages, and other losses like sub-standard quality, are detected and addressed. TPM is often applied when the machinery is costly, which makes it important to bring it into action efficiently. Especially in the food industry, TPM became very popular.
Thomas van de Kracht leads the world class manufacturing program in the Unilever margarine factory in Rotterdam, The Netherlands.
‘TPM ensures that your production lines run stabile, by defining the basis conditions for that’, says Van de Kracht. ‘And by applying 5S, a method to create structured and clean workstations, you will get neatness. Combining these two approaches will make your production processes reliable and, maybe even more important, predictable. When you have reached that level, the way is open to increase the flow and reduce stocks with Lean-techniques. In my opinion applying Six Sigma, to reduce variation and strive for perfection, is the third and final step.’
This line of reasoning is not new. It is well-known that applying TPM results in a well-oiled and trouble-free running factory, but this method also has some shortcomings, which particularly become apparent in the long-term. One of the most important risks is local optimization. The machines run like a charm, but stocks pile up between them, because the processes are not geared to one another.
Therefore is it logical that the next step would be to apply Lean techniques, to improve the logistics by making the processes flow. Next, Six Sigma can be added to strive for a constant and thus always perfect quality.
The Unilever margarine factory in Rotterdam. The upper part (with the logo) is the office of Unilever in the Netherlands (photo Unilever)
WCM old style
It is also possible to start with Lean or Six Sigma. However, to become
‘world class’, in the end you have to develop a customized improvement system, containing components from all three improvement methods: TPM, Lean, and Six Sigma. The definition of World Class Manufacturing (WCM), in this article denoted as WCM old style, is however quite vague:
WCM = TPM + Lean + Six Sigma
Perhaps you noticed that Quick Response Manufacturing (QRM) and the Theory of Constraints (TOC) are not mentioned. Within the context of this article, those methods are seen as supplements or alternatives for Lean. So, WCM aims to improve simultaneously the productivity (with TPM), the logistics (with Lean, QRM and/or TOC), and the quality (with Six Sigma).
Companies claiming they apply WCM - examples can be found via the main menu ‘WorldClass’ on this site - usually started with TPM, ór Lean, ór Six Sigma. Then, after some time, tools of the other two improvement methods were added.
Until recently WCM within Unilever met this description, with in their case TPM as foundation. In the box below, the history and development of TPM and WCM at Unilever during the last 20 years is sketched.
20 years TPM within Unilever
Unilever started already 1996, so almost 20 years ago (!), with the application of Total Productive Maintenance (TPM). Productivity gains of 20 to 30% were no exception during the first ten years. An important mile stone was 2005. That year, two factories received from the Japan Institute of Plant Maintenance (JIPM) an award for world class achievement. This is the highest level you can attain, regarding TPM.
Professor Shirose, one of the founders of JIPM, tastes Unox pea soup
1) Note that the term World Class Manufacturing was not invented by Unilever. WCM has been in use since the 1980s to describe mixed approaches to continuous improvement.
From that moment on, there is some struggle with the strict JIPM-rules. A large company like Unilever clearly benefits from the structure and control through TPM. However, increasingly the question is raised whether TPM, and the related awards, have not become a goal in itself, instead of only a means to improve. Moreover, the need arises to broaden the program, with as most important aim to prevent local optimization, and to make it possible to improve cross-departmental on a supply-chain level.
Already in 2007 the term World Class Manufacturing (WCM) is coined within Unilever1, to indicate that striving for flow (Lean) and the application of Statistical Process Control (Six Sigma) become part of the TPM program. To this end, additional TPM management pillars were added, like customer excellence. Besides this, OEE-measurements for production lines as a whole, and even for complete factories, were introduced. Starting in 2012, Unilever also pioneered with improvement projects coming under more than one TPM pillar, with the aim to make improvement projects crossing departments possible.
Shortly after this, it was decided to implement Yamashima’s WCM-program, as described in the main article. This new approach (1) improves the focus, (2) integrates TPM, Lean and Six Sigma within one program, and (3) increases the attention for human and organizational factors.
During Unilever's long journey from TPM to WCM old style, and finally to WCM new style, one thing remained unchanged: the main goal is to systematically identify and eliminate losses, and to make every employee an increasingly better problem solver.
Starting with TPM, Lean or Six Sigma, and adding the other two methods later, has the advantage that it is clear what is going to happen. Therefore, this probably remains the best approach within companies that start with process improvement.
Nevertheless this add-on approach is somewhat artificial. It is like filling gaps in your original ‘house of improvement’, by adding outbuildings. When TPM is your foundation, this happens almost literally. This is because TPM has areas of special attention, management pillars, like autonomous maintenance and quality. Lean is then for example added by introducing an extra flow-pillar. A similar option is a pillar for customer excellence, this was one of the supplemental pillars Unilever used..
Since 2008, Unilever also produces peanut butter in their factory in Rotterdam (photo Jaap van Ede)
If the foundation of your 'house of improvement' is TPM, an alternative approach has recently become available. In that case no extra pillars are added, but the foundation itself is rebuild. As far as I am concerned, you could call this approach TPM 2.0. In 2005 Hajime Yamashima, professor emeritus at the university of Kyoto and after that consultant, started to develop this WCM new style. For the first time this is WCM formulated as a program, which you can implement step by step.
Launching customer and co-developer was Fiat. They saw this WCM new style primarily as a means to preserve employment in Italy. This explains the emphasis on cost savings, see below.
TPM had zero breakdowns as goal, and Yamashima added zero stock (from Lean and just-in-time production) and zero defects (from Six Sigma). This is called the zero optimum concept. At the same time, production costs as low as possible are pursued. To this end, the problem-solving skills of all the employees must be developed and utilized (Lean). Besides this, the approach is data-driven (Six Sigma), with a lot of attention paid to standard procedures and continuously improving those (TPM).
WCM new style, as developed by Yamashima, defines so-called ‘technical pillars’. The name of these pillars is roughly the same as the management pillars within TPM. Another similarity is that each pillar is managed by one person. However, the content of the pillars is different. Lean and Six Sigma are already ingrained in them, so there is no need for extra pillars later.
Like in TPM, the pillars are implemented step by step. ‘There are seven steps for each pillar’, Van de Kracht explains. ‘Gradually these give a shift from a reactive to a preventive attitude, and finally to a proactive attitude of the employees. Besides this, the order TPM, then Lean and finally Six Sigma returns. Note however that this now relates to building up a pillar, not to the addition of separate pillars as with WCM old style. The first two building steps create order, by defining and controlling basic (production) conditions. Next, the focus shifts to the optimization of value streams. Finally, in step 6 and 7, perfection is pursued.
Customer comes first
Not only the content of the pillars is different, there are also more of them. These extra pillars do not serve to add Lean and Six Sigma to the program, but they do emphasize
special aspects. New is for example a pillar for logistics and customer service, to emphasize that the customer comes first (as within Lean). There is also a pillar for environment and energy, to put sustainability on the agenda. And there is a separate pillar for talent development, to indicate how important the development of problem-solving skills is (Lean). Finally, there is a coordinating, strategic pillar, which resembles Hoshin Kanri (compass for improvement) within Lean. This pillar is called cost deployment, see below.
The way Yamashima remodeled WCM within Fiat was noticed by Unilever. ‘Our chief supply chain officer, Pier-Luigi Sigismondi, is an Italian. He liked the new approach. That way, we thought of implementing WCM new style also within our factories. We started to do so in 2012.’
Professor Hajime Yamashima (on the right), founder of WCM new style, visits the margarine factory in Rotterdam (photo Unilever)
Yamashima’s program seems just right for Unilever, since it gives them the opportunity to grow further with WCM, while preserving all the good things of their solid TMP base.
During Unilever’s long history of applying TPM, on one hand there is a need for a structured and detailed improvement, while on the other hand local optimization has to be prevented, which requires broadening of the program. WCM new style not only solves this conflict, it also adds a strategic element. From now on, improvement actions are not done just like that. Improvement projects that potentially save the most euro’s get priority. Yamashima seems to have adopted this from Six Sigma. That method also estimates beforehand how much money can be saved when a particular improvement project is executed.
The project planning cycle of WCM new style is complex, and is managed within the overarching cost deployment pillar. Filling no less than six matrices, A until F, leads to the selection of the most promising improvement projects.
‘This is repeated every six months, by a team which besides me includes among others our operations, our maintenance and our finance manager’, Van de Kracht explains. ‘At the moment, the whole process goes via Excel spreadsheets. However, we want to automate the data gathering about the actual losses, and converting those losses to potential improvements in Euro's. In the end, we intend to measure in realtime losses of time, labor, materials and energy. That would make it possible to repeat the process of selecting and prioritizing improvement projects much more often.’
The mission of this website: Inspire to create flow in business processes!
Three advantages of free registration
- You will to be able to read all our articles, with many process improvement hints
- You will receive our newsletter every 3 to 4 months, with summaries of all new articles and case studies
(and we promise to send you no other e-mail)
- You will be able to post on our forum
Matrix A: Where (are the losses)
The planning cycle begins with filling matrix A. This matrix indicates where the losses occur. Matrix A contains actual data about time and material losses, energy losses and wasted labor time. ‘Input for this matrix are among others Operational Equipment Effectiveness data, kept by the operators at their production lines’
The team which makes the matrices classifies the losses in three categories, increasing in importance. The first few times matrix A is filled, this classification is done on feelings. Later, it becomes more rational. Only the losses that get a weight of three, are passed on to the next step, matrix B. ‘Converting losses to Euro's is very complicated. Therefore we want to do that only for losses which probably are important.’
Matrix B: Why (are there losses)
For the same reason, another filtering takes place. ‘This happens inside matrix B, which is all about why the losses occur. We make a distinction between causal and resultant losses.’
To explain this, Van de Kracht gives an example. ‘A change-over is a cause of losses. Resultant losses are then for example extra energy consumption, and start-up losses during the next production batch.’
Matrix C: Costs (of the losses)
Only causal losses are transferred to matrix C. In this matrix it is calculated what those losses cost. This is the most daring step of cost deployment, which can be literally translated as ‘forming a battle array on the basis of costs’.
Losses expressed for example in kilowatt hours, kilograms raw materials, waiting times or even meters that people have to walk, are converted to the amount of Euro's that would be saved if a particular loss would not occur anymore. ‘Only then it becomes possible to determine objectively the priority of addressing such a loss.’
One of the converting techniques is Activity Based Costing (ABC). This method attributes costs to one particular activity. However, this raises questions. For example, ABC often indicates that more change-overs, thus working with smaller batches as prescribed by Lean, increases the costs. However, this often is not the case, because cost reduction as a result of having smaller amounts of stocks in other departments outweighs this. ‘We are aware of this issue. In the end, the point is what influence a certain loss has on our sales, and on the costs overall. When cross-departmental effects come into play, we develop methods to compensate for that. In the example you mentioned you could think of calculating the economical most profitable batch size.’
Matrix D and E: Investments (needed to address the losses)
In matrix D knowledge becomes the central point. For each potential improvement project it is determined what boundary conditions should be fulfilled to carry it out. Think of people, tools and training. Next, the necessary investments are weighed against the benefits. Then, finally, the most promising improvement projects are selected on the bases of their so-called ICE-number, which represents the importance, cost and easiness.
In a separate step, projects focusing on sustainability, safety and/or quality are added. ‘On such things you cannot always put a price tag, but you want to be able to focus on it. Via their ICE-number special projects can be placed higher on the priority list.’
Matrix F: Cockpit (to monitor the progress)
Each improvement project that is selected for execution,
is linked to a measurable goal. ‘Suppose you have a number of loss hours as a result of short stoppages of a production line. Then a reasonable aim could be to reduce this amount by 50%.’
The selected projects are transferred to Matrix F, which is used as a cockpit, to track the progress.
To complete this story about filling the matrices: Besides these six matrices, there is yet another matrix G. This seventh matrix contains the expected and the realized savings, and is used for budgeting in the next year.
Determining the priority of the improvement projects with the Cost Deployment process, already led to remarkable insights. For an example, see the box below!
Cost deployment reveals the impact of material loss
In the main article it is described how Unilever determines which losses should be addressed first. This method is called cost deployment. Literally this means forming a battle array of improvement projects, based on the expected profits. Doing this is however a very complex process, so it is an appropriate question whether it is worth the effort. Has it already led to the selection of improvement projects, that otherwise would have been overlooked?
‘Sure’, says Thomas van de Kracht. ‘Let me give an example. Product changeovers are an important cause of losses. To reduce these, we used to apply SMED, Single Minute Exchange of Dies. This method decreases the time you need per change-over. The cost deployment process indicated however that time-related losses, in the form of labor time and machine time, only contribute 20% towards the total costs.’
The remaining 80% of the loss is a result of inferior quality after a change-over, as a result of mixing with remains of the preceding product. ‘When we discovered that, we started to focus on reducing those quality-related losses. For example by alternating our recipes smarter, and by adjusting the design of the machines to reduce the amount of back mixing. That way, we saved 150.000 euro during the first four months of this year alone. If we would have continued to focus only on faster change-overs in the same period, our savings would have been five times less.’
Yet accelerating change-overs remains important too. ‘If you have shorter transition times, your batch sizes can be smaller. This will reduce the amount of stocks in the whole supply chain, and this generates extra savings.’
The cost deployment process selects the big improvement projects, the so-called advanced and major kaizens. ‘Executing those projects resembles a Black Belt project within Six Sigma. But besides these large projects we also have smaller improvement initiatives, in decreasing size these are called standard and quick kaizens. These smaller kaizens are selected and executed by the multidisciplinary teams who operate our production lines.’
Like it is usual within TPM, big improvement projects are assigned to one of the management pillars. However, this is only done in half of the cases, we will return to this later. Each pillar manager not only guides the improvement projects allotted to him or her, but is also busy with developing the pillar itself, which is done via the seven steps mentioned earlier. Interestingly, also these pillar development steps lead to (usually small) kaizens, initiatives to improve the improvement process itself!
Who makes an upstream boat trip along the river 'Nieuwe
Maas', starting at the Erasmus bridge (The 'Swan') in Rotterdam, will soon pass Unilever.
Van de Kracht leads a special WCM-pillar: focused improvement. ‘This pillar does not focus on one specific area of interest. It is my job to see to it, that people stick to the rules of process improvement. For example, do they search for the root cause when there is a problem. I coach the people in the wright direction.’
In a margarine factory a problem with the seals of the margarine tubs could appear to be a quality-related issue at first. However, later the root cause can turn out to be wrongly executed maintenance, or a machine that is not functioning correct. A few years back, Unilever assigned such problems to multiple TPM-pillars. Within WCM new style, there is a much more elegant solution: Problems that need a multidisciplinary approach are assigned to the focused improvement pillar. No less than 50% of the big improvement projects, selected by cost deployment, meet this fate. ‘In those cases it is a priori not known in which direction a solution can be found. I have two full time process improvers, a kind of Black Belts, at my disposal. Those Belts can address these problems in the role of project leader.’
A typical example of a focused improvement project is given in a box below this article.
Van de Kracht stresses that Yamashima’s WCM-approach pays much more attention to human behavior than TPM. ‘When there is a problem, we look at the role of the machine, the materials, the method ánd the people. If human behavior is important, we start a human error analysis. This means that we try to explain why the error occurred. Our primary goal is not to change the behavior. It is better to adapt the work environment, to prevent the problem from reoccurring.’
Unilever cooperated for a long time with the Japan Institute of Plant Maintenance (JIPM). This institution is leading when it comes to the definition of TPM. At the time, JIPM-consultants visited Unilever's factories on a regular basis, and JIPM-awards were granted when a certain TPM-level was achieved.
‘Now we made the step towards WCM new style, similar audits are done by professor Yamashima. Every six months he visits eight pilot sites, which act as model factories. The maximum score of an audit is 100 points, and you can get a bronze, silver or gold award.’
Half of the available points is reserved for the development level of the WCM-pillars. Yamashima defined however also ten other pillars, which can be seen as organizational pillars (sometimes these are called management pillars but this would make this article confusing, since that name is also used for TPM-pillars and WCM-pillars). ‘The organizational pillars are very important, these determine the other half of the points you can get during an audit.’
The organizational pillars are not managed by a person. The name of these pillars refers to organizational aspects, which management as a whole must oversee. Examples are management engagement, clear goals, having a route map to develop WCM, availability of highly-qualified personnel for improvement projects, engagement of the organization, and motivation of the operators.
Unilever is in search of other companies that want to think along about the further development of WCM new style, and with the aim to exchange best practices. When you are interested, you can contact Thomas van de Kracht via email.
93% reduction of losses due to incorrectly placed crimp sleeves
Cross-departmental improvement projects within WCM new style come under the Focused Improvement pillar. Thomas van de Kracht gives an example: the reduction of losses caused by incorrectly placed plastic casings (crimp sleeves) around product packings.
‘Our cost deployment analysis of 2014 showed that a new bottling line for liquid margarine at the time was responsible for the biggest losses’, says Thomas van de Kracht. ‘Short stoppages caused by incorrectly placed crimp sleeves was one of the main causes.’
Those short stoppages not only reduced the Overall Equipment Effectiveness with 7,5%, there were also resultant losses, like wasted packaging material. Within Total Productive Maintenance (TPM) this problem probably would have been assigned to the Quality pillar. ‘But we didn't do that, because we had no idea what the root cause of this problem was.’
Instead, a multidisciplinary improvement team was formed, coached by a project leader from the Focused Improvement pillar. ‘In that team participated technicians, a quality-officer and multiple operators. That way, knowledge of two WCM-pillars was brought in, Quality Control and Maintenance. This was supplemented with the root cause analysis skills of the project leader from the Focused Improvement pillar.’
Processing Point Analysis
Production of Croma,
one of the liquid margarines of Unilever
In this case, Processing Point Analysis
(PPA) was applied to find the root causes
of the incorrect placing of the sleeves
. To this end, you start with a description of the problem and the process point
, the place where the sleeves are added to the products. It is determined which interactions between which machine-components critically influence the result. Next, it is described which sub-systems affect the operation of the process point, and which process parameters are influencing this. Finally it is investigated if the most critical machine parts are in their correct state. If not: ask five times why
As main cause for the problems with the sleeves the team found a 2mm eccentrically mounted gear wheel. Besides this, two other machine parts did not meet standard conditions. These three things were corrected by including them in the schedules for professional and autonomous maintenance. The incredible result: a 93% reduction of the losses! ‘The remaining 7% we will address only when a new cost deployment
cycle would indicate that this is a major loss, compared to other losses. At the moment this is not the case.’
Do you need help with the implementation of WorldClass Operations?
Referral to this article on internet?
Use this link: http://www.business-improvement.eu/worldclass/Unilever_World_Class_Manufacturing_Yamashima2.php