Industrial Objectives and Industrial Performance: Concepts and Fuzzy Handling

By Lamia Berrah, Vincent Clivillé and Laurent Foulloy

This book aims to provide a synthesis of work and ideas done by our team over the last fifteen years in the field of information processing for expression of industrial performance. The statement of objectives on the one hand and the calculation of the other performances are discussed, with the search for the explanation of the link between these two basic steps of an industrial improvement. Beyond the synthetic and typological character of this study, the originality of this work lies in the consideration of the temporal dimension of the objectives, and spread on performance expressions. A fuzzy processing and multi-criteria aggregations time information that can be quantitative, qualitative or symbolic are proposed, in line with industrial practice and literature in the field of performance management.

• Language: English
• Publisher: Wiley
• Release date: Feb 14, 2018
• ISBN: 9781119510611

Book preview

1

The Industrial System

1.1. Introduction

Once upon a time there was a system and an actor. The system functioned and evolved within its environment. The actor, responsible for this operation and this evolution, spent their time observing the system, as a whole, as different parts. They attributed objectives to it, planned actions whose implementation they then managed, expressed the level of performance achieved, and started over with their observation cycle.

So, the tale of objectives begins with a relationship between an actor, a system observer and the system in question. The actor observes the system. Arising from this observation, a representative model is born, brought on by the presence of the actor acting for the system’s structure and operation. Intentions then occur to the actor, for all or part of the system. Therefore, we have the system, the actor, the state of the system observed by the actor, and so the actor’s intention acts as the decision-maker for the system or part of the system. In particular, the actor defines the goals and objectives to be achieved by the system (or part of it).

Thus, the notion of objective emerges from the relationship between the actor and the system. This relationship, both objective and subjective, real and tangible, is based on a large number of aspects that are probably interacting with each other. This is why we will borrow systems theory’s principles and language to comprehend this relationship. Flexible and all-encompassing, systems theory will then allow us to identify links between the various aspects of a system, in particular the entities, the finality, the structure and the behavior... and, consequently, the goals and objectives, all this in a given context and for a given observer.

So, let us begin by recalling some elementary principles of systems theory. Placing ourselves in an industrial context, we will then describe, using the systemic language, what we intuitively call the industrial system. By industrial system, we mean all the operations and all the equipment, used in industrial activities¹. The two latter parts of this description will be dedicated to objective-related information and then to objectives themselves. A representation of the emergence process of the objective, as proposed by the systems theory model, will then round off this exploratory chapter.

But before we get to the heart of the matter, let us take ourselves back to January 2009 and pause to look at the story of Mr. C.C., executive of the RB company and newly appointed associate manager for the Hydraulic Cylinder Production line.

1.2. The RB company’s Hydraulic Cylinder Production line

Resulting from the 2001 merger of companies R and B, the RB² company has a Business Turnover of 4.9 billion euros and 26,000 employees, spread across 41 sites around the world. The RB company is the world leader of the industrial automation and mobile application markets. The company produces hydraulic equipment (proportionately 80% of the Business Turnover), pneumatic equipment (proportionately 15%) and linear guidance parts (proportionately 5%). The RB company designs, produces, distributes and carries out maintenance on all its products.

More precisely, at the company’s Belleville site automation components are produced; mainly Cylinders and Distributors. The technologies used are respectively pneumatic and hydraulic. This production therefore includes four product families, in other words Metal air distributors, Plastic air distributors, Pneumatic cylinders and Hydraulic cylinders. Since the site has had ISO 9000 certification for the last 15 years or so, production is organized into processes. Four value-added processes in particular take place on this site, corresponding to the lines allocated to the four product families which are manufactured on it.

In particular, with its 22 machines, six activities and 30 operations, the Hydraulic Cylinder Production line is dedicated to production of Hydraulic Cylinders. With a production volume of 80 units per day and a range of more than 10⁶ possibilities, the Hydraulic Cylinders are produced in very small batches (mean average size of 1.6 cylinders), complying with classification systems which are broken down into 15 to 20 components depending on the options chosen, with a diameter of between 16 mm and 250 mm and stroke lengths ranging from 1mm to 3000 mm. With opening hours of the order of 10 hours per day and a delivery time of around 3 weeks.

In line with the example set by the RB company, the Belleville site is organized in a functional manner. As the company subscribes to continuous improvement philosophy and development, its organization and working methods are revised regularly. Having recently moved from the Methods and Industrialization department to the Production and Continuous Improvements department on this site, Mr. C.C. takes over responsibility for the Hydraulic Cylinder Production line in January 2009, and has invited us along to experience the first 6 months of his new position at his side, time enough to observe him: observing the line, declaring his objectives, and drawing up the results of some of the actions put in place.

More precisely, Mr. C.C. does indeed have ideas about the operation and improvement of performance of the line. However, in order to be able to specify how his plans will be implemented, Mr. C.C. would like to take the time to observe his system and to understand its inner workings. To do this, Mr. C.C. will spend much time, during the last quarter of 2008, in discussion with Mr. M.N., associate manager of the line since it was set up. To this end, Mr. M.N. begins by broaching the subject of the Overall Equipment Effectiveness – OEE, the Non-compliance rate and the Throughput time.

1.2.1. The Overall Equipment Effectiveness – OEE

Classic productivity indicator, the Overall Equipment Effectiveness – OEE was defined in the 1980s in Japan as being associated, on an elementary level, with the productivity of a piece of equipment within the productive system (machine, production cell, line) [MUC 08]. The Overall Equipment Effectiveness – OEE is computed for predetermined amounts of time, generally a day, a week or a month and applies to both a Piece of Equipment and All Equipment in the system.

The Overall Equipment Effectiveness – OEE is computed as a ratio between the useful time and the Planned production time associated with, respectively, the Piece of Equipment or All Equipment under consideration. The Planned production time is obtained from the Open time of the productive system, from which all the planned stops within the observation period have been removed. The Useful time is computed from the Planned production time by cutting out, this time around, all the unplanned stops (unplanned stops, loss of performance and quality losses) as shown in Figure 1.1, extracted from the standard NF-E60 182 [AFN 02]. To a great extent now standardized, computation of the Overall Equipment Effectiveness – OEE is therefore based on a generic model which identifies all the related types of planned stops and unplanned stops, for the part of the system under observation.

Figure 1.1. Details of the time periods used to compute the Overall Equipment Effectiveness – OEE (inspired from [AFN 02])

For the Hydraulic Cylinder Production line, the Planned production time is known and the unplanned stops are standardized. The latter are manually recorded daily, by staff. The Overall Equipment Effectiveness – OEE is computed weekly. A 65.0% value of this rate represents the expected improvement of the line.

1.2.2. The Non-compliance rate

Intuitive, the Non-compliance rate relates to the compliance of Manufactured products. This rate is an overall computation, on the basis of the ratio between the Quantity of products affected by a compliance problem (i.e. some kind of non-compliance) and the Produced quantity [WEB 12].

As soon as a compliance problem is detected on the Hydraulic Cylinder Production line, the staff member – detector – inputs it manually. Given the line production data, more than 1500 articles are likely to pose a compliance problem, each day. The line’s Non-compliance rate is given to never surpass 1.20%.

1.2.3. The Throughput time

The Throughput time can be defined as follows: the amount of time required for a product to pass through a manufacturing process, thereby being converted from raw materials into finished goods [BRA 14]. Computation of the Throughput time is based on observation of both the value-added time corresponding to line activities and the no value-added time encompassing waiting time, transport and product storage. More specifically, in companies using discontinuous production processes, value-added operations on products generally represent a very low proportion of the time spent by the products on the production lines. Most of the time, the product waits in fact for the whole batch to be finished, for transport to another machine, for a compliance control check... This relationship between value-added time and waiting time can be of the order of 1/10000. In companies with ‘just-in-time production’ this relationship is of the order of 1/100 and, in the best case scenario, of the order of 1/10* [MAR 13].

Computation of the Throughput time for the Hydraulic Cylinder Production line is based on readings made by company employees, who swipe the barcode of each manufacturing order, respectively before and after each value-added operation. These readings are taken respectively in seconds, minutes, hours or days, depending on the type of operation in question. An arithmetical average of the Throughput times for the various Hydraulic Cylinders produced by the line is then computed, generally for a period of 1 month, which generally represents 1500 to 2000 Hydraulic Cylinders. The nominal value of the Throughput time for the line is 8 days.

1.3. Characterization of the industrial system

Consistent with MRP (Materials Requirement Planning and Manufacturing Resource Planning) philosophies about production planning [ORL 75, VOL 04], the industrial system under consideration can be related as a whole to an "organized collection of means implemented with a view to producing tangible goods or to providing services"*³. This production of goods "is carried out by a series of operations⁴ which consume resources and transform the morphological characteristics of the ‘materials’ or modify their location"* [GIA 03]. The transformation or production is triggered by a customer requirement and the resources used are human resources, tangible resources and information [NIE 07]. This transformation is carried out throughout a lifecycle which includes all activities, from identification of customer requirement up to product supply.

At the time that industrial organizations were emerging, industrial activity was essentially focused on the production and its manufacturing factories [BAT 94]. At the instigation of the automobile industry, Taylorian production dealt with men and machines. The process was simple and repetitive, analytically checked. But post-war changes, as much in economic terms as technological ones, had a significant impact on industrial companies, turning this activity into a complex object, and putting many different aspects in interaction with each other [GIA 03, DOU 97], therefore requiring control systems to be put in place. Now associated just as much with the production of goods as with the production of services, and dealing with more or less complex products, industrial activity has widened its perimeter to activities before and after production, encompassing design, industrialization, and logistics as well as support activities such as the compliance, the information system... [BRI 01, BAG 13]. A progressive transition took place moving from the notion of production system/factory/plant to the notion of industrial system.

Progressively, industrial activity has therefore gone from a single supervised process to a collection of control processes. The deserved title of system has been bestowed upon it due to its complexity.

1.3.1. General comments about systems theory

Intuitive, the notion of system has always been used, as soon as a certain complexity was inherent to comprehension of the system. Systems theory – or systemics – emerged around 1945 in the instigation of biological sciences. Construction of the theory is based around the search for analogies between artificial systems and living systems. The purpose of systems theory has therefore been to provide a tool for representation and comprehension of more and more complex objects or phenomena – living or artificial.

Systems theory has caught the eye of various disciplines other than biology. These, information technology [WIE 48], economics [BOU 56, SIM 69], cybernetics [ASH 56], sociology [MOR 77], have all had to carry out Cartesian analysis of phenomena observed from a holistic point of view. Having unanimously approved the complex character of a system, each of these trains of thought has shed some light on this notion, corresponding to their own particular point of view.

A large number of definitions have therefore been proposed for a system (from the Greek word systéma meaning organized set). Even at the beginning of the last century, in 1913, the linguist F. de Saussure saw a system as an organized whole, made of connected parts which can only be defined by referring to each other and as a function of their place in the whole* [SAU 95]. The biologist L.V. Bertalanfy identifies a system as a complex of elements standing in interaction [BER 68]. In a similar way, the information technology community defines a system as a combination of interacting elements organized to achieve one or more declared purposes*. [ISO 15]. This group forms the structure of the system as described by J.L. Le Moigne [LEM 94]. Considering the system from the point of view of its behavior, the sociologist E. Morin states that a system works and transforms itself for a number of finalities* [MOR 77]. Instead of the plural used by that author concerning the concept of finality of the system, we prefer to use a singular, placing ourselves at a more general level, and proposing to identify the finality of any system by its durability. On the contrary, to be guaranteed, this finality of durability will need to be specified, in the form of goals, then objectives and actions….

In this respect, H. Mintzberg distinguishes the mission goals and the system goals [MIN 96]. The mission goals guarantee the finality of the system. They are based on the system’s mission, of which one aspect consists of fulfilling the expectations of the system’s environment. The system goals are the basis for system operation, they are often connected to satisfaction of the internal constraints required to maintain the system in a condition that allows it to operate and fulfill the mission goals. In relation to previous categories of goals, systems theory experts also talk about exogenous objectives and endogenous objectives, referring to the impact, whether external or internal to the system, of the actions associated with these objectives.

Along the same lines, although resisting the will to apply a typology, we could also evoke a different property of the objectives and goals of a system, that which is related to their, respectively, structural or conjunctural nature. Indeed, one or the other can be recurrent and repetitive, just as they could be linked to some kind of request or to the occurrence of an unplanned event [CLI 04]. Figure 1.2 shows this first set of terminology about objectives.

Figure 1.2. Goals and objectives

Operation and transformation of the system will be affected naturally by the objective. Whatever the case and beyond this initial characterization, goals and objectives, motivational by nature, have the common special feature of being part of performance improvement of the system.

Consequently, operation of the system is based on its activity, its dynamic [FOR 61]. Its transformation denotes its evolution [LEM 94]. The system operates to reach the mission goals. Reaching these mission goals is simultaneously associated with reaching system goals. In addition, if the system cannot reach its mission goals or its system goals, an inadequacy or an improvement opportunity is detected. The system then transforms, as does simultaneously its operation. As soon as the system goals begin once more to be reached, and consequently the mission goals, the system stops transforming [CLI 04]. To give the whole picture, let us observe that an improvement can also be undertaken even if the mission goals or the system goals have been reached. Effectively, the system could naturally be compelled to transform itself and the mission goals and system goals could be driven to evolve due to this, in a will for pure improvement, however, without any inadequacy being detected.

While on this subject, let us specify that, depending on the system in question, this operation, as well as this transformation, can be more or less automatic. The latter can therefore be controlled, to a greater or lesser extent, by human beings, i.e. the actor, system observer in our case.

The epistemologist J.L. Le Moigne defines a system as a representation of an active phenomenon, comprehended as identifiable by its projects, in an active environment, in which it functions and transforms teleologically⁵ [LEM 90]. All data and events which can affect the system’s behavior will be dictated by the system’s environment. Focusing on industrial organizations, H. Mintzberg specifies that the behavior of the industrial system is above all related to its operation, i.e: implementation by means of sociotechnical resources of a set of activities to generate value in the form of goods, in compliance with the mission goals* [MIN 96].

All of this thinking has led to the development of systems theory modeling methods [LEM 90]. The issue with systems theory modeling of a reality is then the comprehension of this reality by means of what it does and how it does it⁶ [BOU 56, LEM 90]. Often seen as being contradictory to analytical modeling, systems theory modeling considers the system as a whole, indissociable, whereas the analytical psyche believes that the system can be reduced to the sum of its parts.

Thus, systems theory can be seen as a flexible and all-encompassing modeling framework. To this effect, Figure 1.3 summarizes the paradigm (all principles), in accordance with the mnemonic description, repeated below, of J.L. Le Moigne in his book General systems theory, theory of modeling:

– "something (anything, presumed to be identifiable),

– which within something (environment),

– for something (finality or project),

– does something (activity = operation),

– by something (structure = stable form),

– which transforms with time (evolution)."

Figure 1.3. The systems theory paradigm [LEM 94, p. 82]

Approaches to systems theory modeling can vary from one discipline to another. However, they all characterize the system on the basis of the definition of its structure and its behavior, and rely on certain fundamental principles to do this [DON 02]. Among these principles, we will retain:

– the role of the system observer, on the one hand;

– the levels of abstraction at which the observer observes the system, on the other hand.

Before we continue our characterization of the system with respect to structure and behavior, let us take the time to delve further into the two fundamental principles which are of interest to us. Let us recall that this interest comes from the fact that our aim in this book is to look at outlining objectives, where the relationship to the observer who declares them, as a function of the abstraction that the latter has of the system, is established quite naturally.

1.3.2. The role of the observer

In the systems theory model, a system is only known if it is associated with an observer. In other words, a system can exist but only becomes a known object and/or an object of interest if it starts to be observed. To demonstrate the importance of this notion of observation, let us borrow, for a while, a portion of The Little Prince’s conversation during one of his encounters: We only really know things that we can tame, said the fox. ... What does ‘tame’ mean? – It’s something too often forgotten, said the fox. It means ‘create a bond …. ’ …. If you tame me, we will need each other. In my eyes you will be unique the world over. In yours I will be unique the world over⁷.

When we say observer, we mean the actor who steers the system, i.e. observes and operates the system (decides and takes action) [DON 02]. And to illustrate this notion of steering a system, let us ask the Little Prince about his view of steering his rose. Of course, my own rose, she, would look just like all of the rest of you to an ordinary passer-by. But she alone is more important than all of you put together, because she is the one I have watered. Because she is the one I have sheltered under glass. Because she is the one I have protected with windbreaks. Because she is the one I have picked caterpillars off to kill (except for two or three for the butterflies). Because she is the one I have listened to when she complains, or boasts, or sometimes just stays silent. Because she is my rose⁸.

The observer will then put forward a system model. Intervention by an observer who sees themselves as a system will express the intentionality of knowledge [LEM 94]. The system model therefore depends on the observer’s point of view, i.e. from the point of view that the observer wishes to use in order to model the system. People see stars in different ways. For some, who travel, the stars are their guides. For others, they are nothing more than little dots of light. For others still, who are wise, they are problems. For my businessman, they were gold⁹.

To model the system, the observer adopts a particular point of view, taking into consideration a certain number of parameters. Other than objective measurements, a part is also played by how well they know their system, its finality and its environment, as well as their understanding of the interactions between the two, all of which we can call the observer’s expertise [END 99, MON 15]. On the contrary, this point of view can include the observer’s intention for the system they are observing, with respect to the requirements they notice or the desires they have for the system. Generally, we can consider that this point of view is affected by what is known in more broad terms as the observer’s attitude [SLO 02]. The subjective nature – which we see again in the following chapters – of this