Being the best, By checking the rest

(1)

Being the best,

By checking the rest

Centralized benchmarking of manufacturing within Akzo Nobel

Appendices

A study ordered by and performed at:

Akzo Nobel Chemicals Manufacturing Support

As a master thesis for the completion of the study of:

Industrial Engineering & Management

At the:

Faculty of management and organization, University of Groningen

Deventer, 6 December 2005

Author: Daniël Hubbeling (s1063197)

Supervisors University of Groningen: Dr. E.O. de Brock Prof. Dr. J. Wijngaard

Supervisor Akzo Nobel: J.P.M. van Aggelen

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying recording or otherwise without the prior written

(2)

(3)

Appendix A - Approaches to maintenance

(Source: ManSup)

When looking at maintenance three elementary types can be distinguished. In practice, all implementations of maintenance within companies will consist of a mix of these three elementary types.

Failure dependent maintenance

Failure Dependent Maintenance implies that maintenance of installations takes place at the moment when failures occur. Failures occur mainly due to errors of the operator, rupture or wear. This type of maintenance can mean an unplanned downtime of the installation causing a loss of production. In practice, the ability to improvise of the operator and/or mechanic is very important to limit this loss of production to a minimum. The focus of this kind of maintenance does not lie with the cost, any form of cost control is often lacking entirely.

Maintenance strategies mainly based on failure dependent maintenance are often seen in companies with relative simple hardware who only work in day shifts. When the consequences of possible breakdowns are relatively low and the failure interval is long, this method is suitable for organizing maintenance.

In practice, this kind of maintenance is used in (nearly) all manufacturing processes because it is virtually impossible to prevent failures resulting in unplanned downtime completely.

Preventive maintenance

A second form of maintenance is preventive maintenance or usage dependent maintenance. This implies that maintenance occurs at a predetermined time. The actual moment of maintenance can be planned at a fixed moment in time or at a fixed point on another scale on which the use of installations can be measured, like the total amount of produced goods or the distance driven by a car.

This kind of maintenance can cause certain parts of installations to be replaced while they have not yet reached the end of their technical life span.

Predictive maintenance

A third form of maintenance is predictive or condition-based maintenance. With this kind of maintenance, the inspection of installations plays a big part. It implies that maintenance is carried out based on findings of planned inspections. These inspections are also used to estimate the remaining life span of certain parts. Knowledge of the entire installation and/or certain parts is very important for this kind of maintenance. A condition for this kind of maintenance is that the state of (a part of) the installation can be somehow measured and boundaries within which normal operation is possible can be given.

A development in this kind of maintenance is the so-called condition monitoring in which (electronic) measuring devices are used to determine the condition of parts of the installation. This development makes it possible to monitor parts without actually stopping production and disassembling installations.

(5)

Reliability Centered Maintenance (RCM)

(From: Productivity Futures, a Productivity Portal maintained by the Institute of Management Services and the World Confederation of Productivity Science, http://www.lmu.ac.uk/lis/imgtserv/tools/rcm.htm)

Reliability centered maintenance (RCM) can be defined as a systematic process of preserving a system's function by selecting and applying effective preventive maintenance tasks. The main difference from most approaches to preventive maintenance, however, is that it focuses on function rather than on equipment.

The RCM approach was developed in the late 1960’s and early 1970’s when the increasing complexity of systems, and consequent increasing size of the preventive maintenance task, forced a rethink of maintenance policies among manufacturers and operators of large passenger aircraft.

The principles that define and characterize RCM are:

• A focus on the preservation of system function;

• The identification of specific types of failure to define loss of function or functional failure;

• The prioritization of the importance of the types of failure, because not all functions or functional failures are equally important;

• The identification of effective and applicable preventive maintenance tasks for the appropriate types of failure. Applicable means that the task will prevent, relieve, detect the onset of, or discover, the failure mode. Effective means that among competing candidates the selected preventive maintenance task is the most cost-effective option.

These principles, in turn, can be implemented in a seven-step process:

1. The objectives of maintenance with respect to any particular item/asset are defined by the functions of the asset and its associated desired performance standards;

2. Functional failure, meaning the inability of an item/asset to meet a desired standard of performance, is identified. This can only be identified after the functions and performance standards of the asset have been defined;

3. Types of failure, which are reasonably likely to cause loss of each function of the asset, are identified;

4. Failure effects, describing what will happen if any of the types of failure occur, are documented;

5. Failure consequences are quantified to identify the criticality of failure. RCM not only recognizes the importance of the failure consequences but also classifies these into four groups: hidden failure, safety and environmental, operational and non-operational;

6. Functions, functional failures, types of failure and criticality are analyzed to identify opportunities for improving performance and/or safety;

7. Preventive tasks are established. These may be one of three types: scheduled on-condition tasks, which employ condition-based or predictive maintenance, scheduled restoration and scheduled discard tasks.

(6)

Although one of the main objectives of RCM is to reduce the total costs associated with system failure and downtime, evaluating the returns from an RCM program solely by measuring its impact on costs hides many other less tangible benefits. Typically, these additional benefits fall into the following areas:

1. Improving system availability; 2. Optimizing spare parts inventory;

3. Identification of component failure significance; 4. Identification of hidden types of failure;

5. Discovering significant, and previously unknown, failure scenarios;

6. Providing training opportunities for system engineers and operations personnel; 7. Identification of areas for potential design enhancement;

8. Providing a detailed review of plant documentation, and improving where needed.

Total Productive Maintenance (TPM)

(From: Productivity Futures, a Productivity Portal maintained by the Institute of Management Services and the World Confederation of Productivity Science, http://www.lmu.ac.uk/lis/imgtserv/tools/tpm.htm)

Total Productive Maintenance (TPM) can be defined as the systematic execution of maintenance by all employees through small group activities. TPM has two goals: zero breakdowns and zero defects. This obviously improves equipment efficiency rates and reduces costs. It also minimizes inventory costs associated with spare parts.

TPM is often defined as productive maintenance involving total participation. Many organizations misconstrue this to imply that only shop floor staff needs be involved. However, TPM should be implemented on a company-wide basis to result in the sought after improvements of efficiency and cost reductions.

TPM aims to establish good maintenance practice through the pursuit of five goals:

1. Improvement of equipment effectiveness: examine the effectiveness of facilities by identifying and examining all losses that occur - downtime losses, speed losses and defect losses;

2. Achievement of autonomous maintenance: allow the people who operate equipment to take responsibility for, at least some, of the maintenance tasks. This can be at :

• The repair level, where staff carry out instructions as a response to a problem;

• The prevention level, where staff take pro-active action to prevent foreseen problems;

• The improvement level, where staff not only takes corrective action but also propose improvements to prevent recurrence.

3. Plan maintenance: make sure there is a systematic approach to all maintenance activities. This involves the identification of the nature and level of preventive maintenance required for each piece of equipment, the creation of standards for condition-based maintenance, and the setting of respective responsibilities for operating and maintenance staff. The respective roles of ‘operating’ and ‘maintenance’ staff are seen as being distinct. Maintenance staff is seen as developing preventive actions and general breakdown services, whereas operating

(7)

staff take on the ‘ownership’ of the facilities and their general care. Maintenance staff typically moves to a more facilitating and supporting role where they are responsible for the training of operators, problem diagnosis, and devising and assessing maintenance practice.

4. Train all staff in relevant maintenance skills: the defined responsibilities of operating and maintenance staff require that each employee has all the necessary skills to carry out these roles. TPM places a heavy emphasis on appropriate and continuous training.

5. Achieve early equipment management: the aim is to move towards zero maintenance through ‘Maintenance Prevention’ (MP). MP involves considering failure causes and the maintainability of equipment during its design stage, its manufacture, its installation, and its commissioning. As part of the overall process, TPM attempts to track all potential maintenance problems back to their root cause so that they can be eliminated at the earliest point in the overall design, manufacture and deployment process.

TPM is effective in eliminating losses in the following areas:

1. Downtime from breakdown and changeover times;

2. Speed losses, when equipment fails to operate at its optimum speed;

3. Idling and minor stoppages due to the abnormal operation of sensors, blockage of work on chutes, etc.;

4. Process defects due to scrap and quality defects to be repaired;

(8)

Appendix B - System of profound knowledge

The successful application of a continuous form of benchmarking, as described in this report, will not only depend on the tools that are made available. The prevailing style of management within that organization will also play an important role. The goal will eventually be to come to a culture that is focused on continuous improvement of the performance of the entire organization rather than on all individual parts of an organization focusing on maximizing their own production, sales, profit or any other competitive measure.

This idea was formulated by Dr. Deming in his System of Profound Knowledge (Deming, 1994). The System of Profound Knowledge is meant to provide an outside view on organizations and consists of four elements that are all related to each other. The four elements are: appreciation for a system, knowledge about variation, theory of knowledge and theory of psychology.

The purpose of this appendix is to reflect on the description of the benchmarking project that is designed in this study, both regarding the theory on benchmarking and the organization/process around it, from a somewhat higher perspective.

Appreciation for a system

Deming defines a system as a network of interdependent components that work together to try to accomplish the aim of the system. The greater the interdependence between components, the greater the need will be for communication and cooperation between them. As a consequence the need for overall management will be greater too. (Deming)

An example to illustrate the interdependence of components Deming gives is that of an orchestra. The members of an orchestra do not aim to play solos as ‘prima donnas’ but they aim to support each other as the way they will be judged by listeners is how they work together in their performance. The conductor, as manager of the orchestra, tries to improve the cooperation of the players. The level of interdependence between components is shown in figure A-1.

Figure B-0-1: Interdependence, from low to high

When looking at the context for this benchmarking project two distinct systems, each with their own goal, can be identified.

1. Plant Organization

In the plant organization the actual manufacturing of chemicals takes place. In this system different disciplines work together to produce their products in an effective and

Level of interdependence

High Low

Bowling team Orchestra Business

(9)

efficient way and thus make profits for their Business Units and Akzo Nobel as a whole. The three disciplines technology, production and maintenance have been selected as the different components with which this system is described in this study. However, other elements, like procurement and logistics, have a role in this system too.

2. Akzo Nobel organization

The whole Akzo Nobel organization is considered as a system. In this system the different Business Units, as the different components, aim to maximize the profits for Akzo Nobel as a whole. Compared to the first system at plant-level, the interdepedence between the different Business Units is relatively low. The Business Units have in common that they operate plants for the manufacturing of (chemical) products. It makes sense, to say the least, that they share knowledge. A shared benchmarking study, based upon the components of the greater Akzo Nobel system, is one possibility to facilitate this sharing of knowledge.

The notion that it would be counterproductive to try to maximize any aspect of a system without looking at the effects for the system as a whole is also consistent with the application of the Balanced Scorecard methodology as it is described earlier in this report.

Knowledge about variation

The second element of the system of profound knowledge is knowledge about variation. Variation comes from the environment, from processes, from equipment, from materials, and from people. (Deming, 1994)

Variation is inherent to every system. And thus, understanding variation is very important when interpreting any kind of measurement of a process.

When it comes to reacting to outcomes of measurement two common pitfalls can be identified, when variation is taken into account. They are:

• Pitfall 1: React to an outcome as if it came from a special cause, when actually it came from common causes of variation.

• Pitfall 2: Treat an outcome as if it came from common causes of variation, when actually it came from a special cause.

When this is applied to the case of the benchmarking, it essentially means that managers need to understand whether any score they get on any of the performance indicators is the result of common causes or of a special cause. If it is the result of a special cause it may mean that any quantitative comparison of that particular performance indicator in the context of the benchmarking exercise is not valid.

However, any disturbances of the process that serves as this ‘special cause’ is in itself a signal for a possibility for improvement.

(10)

Theory of knowledge

The theory of knowledge helps to understand that management in any form is prediction. Rational prediction requires theory and builds knowledge through systematic revision and extension of theory based on comparison of prediction with observation. (Deming, 1994)

The theory of knowledge is the most difficult part of Deming's system of profound knowledge for most to understand. And as a result it is also an area that often is under appreciated. Some of the difficulty of managing with data can be minimized with an understanding of the theory of knowledge.

Many fail to predict when attempting to experiment and test. Without prediction learning is much less than it would be. When it is understood that management is based on prediction then the impact of all the other three areas of the system of profound knowledge are clearer. By exploring the basis for the prediction that one is making, one must understand the theory they are using to make the prediction.

Most often people fail to develop a theory that allows them to predict (they just act without theory or with an undefined vague idea of what they expect). They fail to predict the results of an experiment and they fail to analyze the results of the experiment. So they fail to learn about the system that they are managing and therefore cannot refine their theory based on their learning. The failure to do these things will lead to very ineffective learning. And without learning any significant improvement should not be expected.

When applied to this benchmarking system, this means that, although the results of the quantitative comparison might be similar to what was expected by managers, these results may prove to be a stimulus to act on any performance gaps that is identified, and improve the knowledge that exists on the reasons for this performance gap’s existence.

In addition to this, the quantitative analysis is essential in pointing to where best practices can be found. Even if all it does is to acknowledge the existence of performance gaps that were already suspected to exist.

Another important thing that can be concluded from this part of the theory is that any quantitative comparison in the context of this benchmark is only useful if the definitions and working methods that are behind the performance indicators are understood by the person analyzing the results. If this is not the case any conclusion that is made based on the results probably is not valid and may lead to implementation of changes to the underlying system that prove to be ineffective.

Theory of psychology

Psychology helps to understand people and circumstances, interaction between people, interaction between a manager and his people, and any system of management. (Deming, 1994)

In order to get good benefits from the benchmarking system as described, the psychology of all people involved shall be observed. This is a complex matter since all individuals react in their own way. The motivation of employees can influence the performance of

(11)

plants significantly. This motivation is influenced by several sources and phenomena, being:

• Intrinsic source of motivation o Self-esteem,

o Desire to learn, o Joy in work,

o Relationship with colleagues, o Family relation

• Extrinsic source of motivation o Salary,

o Bonus, o Ranking • Overjustification

o Awards for achievements which are considered as the normal task, o Awards from somebody who does not have the respect of the receiver The bottom line is that in general individuals are less motivated by money than by getting the occasional “pat on the back”. The benchmarking system and how results are interpreted and used need to be considered in view of the above psychological effects it will have on individuals. Especially the ranking of performance of plants needs to be done without pointing fingers to individuals, either in management or in operational functions.

As stated earlier results of a plant are influenced by variation in the environment, processes, equipment, materials and people. Human nature, however, is to attribute any result to people for the major part. The results are projected mostly on people, which are closest to the problem areas within the organization. The search for the innocent and/or the guilty individuals is in most cases a waste of time and effort, the focus should be in making changes in the system to prevent problems or improve performance in the future.

The final goal for the benchmarking system is to get information from the plants to improve the organization, by implementing changes that improve quality, prevent errors and reduce waste.

In order to have a proper control over the improvement process the actual values (measurements) from the organization need to be a true reflection of reality. In addition to the proper definition of the measured performance indicators it is of equal importance to get correct information. The information, used as input for the benchmarking system, has a great human interpretation influence; hence the human factor needs a lot of attention.

If the results are not used in a proper way individuals will see the benchmarking system as a threat. If this happens there are some pitfalls to observe:

• Individuals will (try to) distort the system, by focusing on one aspect of the system too much and improving this at the expense of the whole system;