||Operator in the role of Sherlock Holmes at Nutricia
By Dr Jaap van Ede, editor-in-chief business-improvement.eu, 22-04-2014. This article was published earlier in the Dutch specialist journal Fluids Processing1
The Overall Equipment Effectiveness (OEE) is defined as the fraction correctly made products relative to the maximum amount possible. By monitoring the OEE during each shift, it is precisely known when disturbances and problems occurred.
Yet, as Nutricia in the Netherlands discovered, the current reporting methods are not adequate enough to determine the root cause of all problems. Inquiring from the operators then takes a lot of time, and the event is not fresh in their mind anymore. Therefore, the operators were recently trained to capture as much information as possible in the case of any significant disturbance. They even secure evidence as if they were detectives!
To this end, within an existing OEE-toolkit, a template was developed. Digital forms help the operators to make good problem descriptions. The forms are framed by Sherlock Holmes-like drawings.
Nutricia in Cuijk in the Netherlands became part of Danone in 2007. Nutricia produces baby food, mainly products in the form of powder which are sold under the brand name Nutrilon. Dissolved these are a substitute for breast milk. Nutricia produces many variants, for example for children who are allergic to food ingredients.
It goes without saying that the quality requirements are high. After all, the scandal with contaminated milk powder from a now bankrupt Chinese company is still fresh in memory.
It is also important to produce efficient, with as few losses as possible. This results in a high price-performance ratio. To detect losses and to prevent those in the future, Nutricia measures the Overall Equipment Effectiveness or OEE. Already for years, this is done with the aid of a software toolkit, originally installed by Blom Consultancy. The data-acquisition goes manually. However, at the moment a promising pilot is done with automatic input.
For each production line, the OEE-toolkit registers the amount of correctly manufactured products relative to the maximum amount possible.
‘When there is a significant disturbance or loss, the operator checks off in a digital list what the problem was. Via a form he or she can add a full explanation when needed’, says Arjan den Hartog, performance manager in the Nutricia-factory in Cuijk. ‘Later, we can use the data to trace the disturbances that cause the largest losses, for example by doing a Pareto-analysis. And we can form multidisciplinary improvement teams, to find and tackle the underlying problems.’
Good chance that you think: Nothing new under the sun, in our own company we are already doing this for years! ‘The difference is in the details’, Den Hartog explains. ‘In the most basic form, as I described above, the system did not function well. The problem descriptions provided by the operators often turned out to be incomplete. Information was missing when we wanted to take action to solve the problem, to prevent it from happening again. Of course, at that moment we could inquire with the operator concerned. However, that is difficult when the work is done in shifts.’
To explain this, Den Hartog first expounds on which levels problems are detected and solved. ‘We have three levels of performance measurement. On an hourly basis, the operators get feedback, for example about the amounts produced versus the planning. That way, they can immediately take action in case of deviations. The second feedback moment is when the new work shift arrives. At that moment, the operators pass on problems and priorities to each other. By this means, a lot of issues are solved immediately.’
The third feedback moment is the daily production meeting per department. ‘During that meeting, all significant disturbances in the past 24 hours pass in review. However, only one operator of the department concerned is present at that moment. Therefore chances are high that an issue comes up, that was not witnessed personally by that person. At that moment we notice the troubles that incomplete problem descriptions bring.’
I ask Den Hartog to clarify this with an example. ‘Imagine that a baby food packing machine got jammed for ten minutes during the morning shift. The operator made a problem description, which says that a mechanic hastened to help and solved the issue. During the afternoon shift a similar problem is reported, and again it was solved by a mechanic. In this case, the following questions will arise during the production meeting the next day: Was it the same problem, or did the second mechanic reinstate the machine? To answer these questions we often had to trace the operators concerned. That takes a lot of time, because we work in shifts. In addition the information isn’t fresh anymore in the operators minds. Besides that, important stuff like damaged packing material is possibly gone.’
Therefore, the key question was: How can we ensure that the operators, when an incident happens, make a problem description that is very accurate. So accurate, that immediate action can be taken when the problem is discussed during the production meeting?
‘The information you need is represented in large part by the acronym 5W2H. This is a well-known aid to describe problems by recording Who, What, When, Where, Why, How and How much. The “How” we extended to “How was the problem solved”. For example, the operator indicates if it is a temporary fix, or that the problem was solved permanently. That addition alone was a very big improvement.’
^ Packaging line for Nutrilon baby food. The operators lay down all information
about production disturbances, like they were detectives!
To help the operators to make quick and accurate problem descriptions, to the OEE-toolkit a form was added with seven questions. Briefly, those questions are:
- Who detected the problem?
- What happened?
- Where in the process or at the object?
- When did the problem occur?
- Why was it a problem?
- How often / how much did it occur?
- How was the problem solved?
Providing only these questions still proved to be insufficient. One example: The When-question could be answered by filling in just the time. Often, it is also important to know if the problem occurred during a changeover, during the startup of production, or in the middle of a production run.
'Therefore we added hints to the questions. These help to lay down all information that could be of interest to find the root cause of the problem. At question 7 it is for example pointed out that it can be important to keep for example damaged packaging material. This way we make our operators behave like detectives, which safeguard as much proof as possible. When a crime is committed, you have a crime scene and you collect information to trace the culprit. In our case we have a “process breakdown scene”. At that location we collect information to find the root cause of the breakdown.’
Working with the forms might have a disadvantage: Every problem becomes unique in the database. That would make it hard to determine which problems occur the most via a Pareto-analysis.
‘That is true. Therefore we continually assess if it is better to add a certain problem to the drop-down list with standard situations. However, having a very long list is cumbersome and paves the way for mistakes. To prevent that we regularly remove items from the list, for example problems that don’t occur anymore.’
When the new system was introduced, initially the operators were afraid that it would increase their workload. ‘We could remove that fear in part, by restraining the use of the forms to medium and large disturbances. Those are defined as lasting longer than five minutes. Besides that, we explained that the system would lead to fewer questions afterwards. Finally, it is important that an operator sees that his or her input actually leads to improvement. At the moment we already provide that kind of feedback. However, I think we should organize it in a more structured way.’
Quite often, people from other Danone factories visit Nutricia. ‘Those guests are impressed, both by the design of our system and the discipline of our operators, who complete the forms. We coach them to do that task increasingly better. When an operator describes things unclearly, we explain why. At the moment, 95% of the problem descriptions is accurate enough to deal with these immediately, I mean without consulting the operator.’
^ Arjan Den Hartog, performance manager at Nutricia: "During production
disturbances we record 5W2H: Who, What, When, Where, Why, How & How Much"
Continuous improvement by way of OEE measurements is part of Total Productive Maintenance (TPM). ‘But that is not the only improvement method we use. Danone, our parent company, has developed an operational excellence program, which is called DaMaWay. This stands for Danone Manufacturing Way. Besides TPM the program also encompasses the application of Lean manufacturing and Six Sigma.’
The system for OEE-measurement, described in this article, is only one of the many tools of the DaMaWay. ‘Another example of a tool, that also has the aim to improve the productivity, are the brainstorm sessions around a particular machine. During such a session a multidisciplinary team gathers, to evaluate the state of affairs with a helicopter view. At first, we do this around the machines that have the biggest downtime. The team also addresses the reduction of the energy consumption and the environmental impact. For that purpose our nature & environment manager is present. In every factory of Danone someone has that role.’
Danone has a global sustainability strategy that is bearing fruit. In 2008 it was decided that the carbon footprint, measured in kg CO2 emitted per kg product over the whole supply chain, should be 30% lower in 2012. To this end a measurement system was built, based on the ERP-system SAP. The goal of 30% lower CO2 emission has already been achieved.
Login to read more!Login
to read more (after login, you will be redirected to the current article)
Some articles like this are (partly) hidden, for example because these were published very recently in a logistic specialist journal.
Three advantages of free registration
- You will to be able to read all our articles and business cases, 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
1) Published in the Dutch journal Fluids Processing april 2013
Do you need help with the implementation of TPM?
Referral to this article on internet?
Use this link: http://www.business-improvement.eu/tpm/Nutricia_Danone_TPM.php