A common concept of activity for production systems ... - Riad Megartsi

makes difficult to be reactive. In order to overcome these problems, the ... design process and do not enable a direct extraction of information for the control.
38KB taille 9 téléchargements 314 vues
A common concept of activity for production systems and design-processes control

MEGARTSI Riad, CAUVIN Aline, AYADI Karim. DIAM (Département de recherche en Informatique, Automatique et Mécatronique)-IUSPIM, Domaine Universitaire de saint Jérôme, Avenue Escadrille Normandie Niémen, 13397 Marseille CEDEX 20, France. Tel : +33 4 91 05 60 21 - Fax : +33 4 91 05 60 33 E-mail : [email protected], [email protected] & [email protected] Abstract - This paper deals with the perspectives of an analogic transfer of the activity concept between production systems and design-processes control. The purpose is to present a common concept which improve the interrelations between the actors of the enterprise. Keywords - Activity, Modelling, Control, Production systems, Product design processes.

1.

Introduction

The present economic situation incites the industrial firms to improve their internal reaction processes to face disturbances in order to stay competitive. Moreover, the participation of several actors from different disciplines makes difficult to be reactive. In order to overcome these problems, the enterprise modelling has become a main preoccupation. In this paper, we present a brief survey of the existing methods and tools in production systems and designprocesses control. Our work proposes to confront these methods and tools, in way to develop a common modelling of the activity concept. 2.

Modelling of the production systems control

The set of decision and information sub-systems constitutes the production control system [Fig1].

Disturbances Information Sub-system

Decision Sub-system

Raw material Components

Finished products

Operating sub-system

Figure 1: The operating, information and decision subsystems of a production system.

The problematics of the production control can be formulated as follows: how can we develop decision support systems which make a group of actors, holding different functions, possible to seek their own objectives in agreement with those of the enterprise? The main methods (CIMOSA, IDEF, GIM, PERA) presented in [6] allow to build models that cover only few sides of the production system. In addition, their objectives are to develop a reference of the production system instead of interesting to the control mechanisms. Moreover, they focalise on the decisions daily taken and do not take into account the occurrence of disturbances.

3.

Modelling of the design processes control

The problematics of the design processes control can be formulated as follows: how can we manage the product knowledge evolution and how to provide a traceability of design choices? How can we represent the decisions taken in design, in order to choose the relevant solutions to react against disturbances? How can we provide a common language for several disciplines actors participating to a design project? The representation of the design processes control is realised through models. These models allow to structure information that follows the product all along his life cycle. All the models presented in [6] do not cover the entire design process and do not enable a direct extraction of information for the control.

4.

Convergence between the two domains of control

formalism based on activity concept taking into account its skills and performance indicators. Initial state

All the former approaches are built on the concept of activity for task control modelling. Each of them presents a specific method regarding the predetermined objectives, in order to represent some views of the production and design processes.

Final state

Control processes

Activities network

The convergence between the control of the production systems and the design-processes is the subject of researches [4], [7] and shows similar principles of control. In a matter of fact, we find in both control cases: -

the use of common notions of planning, reactivity to respond to constraints and disturbances;

Figure 2: Control process representation.

-

the need for a good managing of the quality, and a traceability of the products and their associated processes;

-

the optimisation of the allocation human resources;

Complementary, we propose a common concept of activity [Fig 3], in order to build a modelling approach independent from the studied process (design or production). It is characterised by the following properties:

-

the need for coordination, communication and partnership between the different actors from different disciplines;

-

the need for the know-how capitalisation, in order to re-use it in following projects;

-

the necessity to better manage the skills of the human resources and the mechanisms used for problem solving;

From the study of the problematics of the control modelling, two interrogations arise: which modelling entities are common to the domains of the production systems and design processes control? How to adapt these entities to improve communication between actors from different culture? In our opinion, a common concept of activity can answer these questions. 5.

The common concept of activity

We aim to capture the description of real processes the more complete as possible as existing in an enterprise. Such a description constitutes an analysis tool. It will be a support to make easier interrelations between actors. Particularly, it improves control processes used to face disturbances [Fig 2]. In that purpose, we consider analogic transfer of the activity concept between the domain of control production and design process. Our common concept of activity is based on IDEF3 formalism [5] and its extension IDEF3x, developed by the members of the Acnos project [1].

-

it provides a unique representation of the processes composed by activities of the production management and design. Therefore, it simplifies the communication between actors from different discipline;

-

every activity is identified by its name, a code, its duration and a verb which describes its objective;

-

the common concept of activity identifies the required human skills;

-

it characterises the required and effective performance indicators (cost, time and quality) to express the constraints and the objectives of every process and activity. required effective

Performance Indicators

cost quality time Verb

Type (S, I, E)

Duration

Code

Human resources

Available human resource

Technical resources

Code

Skill

Required skill of human resource

Figure 3: the common concept of activity. The IDEF3 method allows to capture descriptions of sequences of activities and to model the precedence and causality relations between processes and events in their environment. IDEF3x proposes a representation

This representation distinguishes three types of activity: execution, information and selection characterized activities [2]:

- Information-characterized activity: it consists in picking up and transmitting information without transformation. - Selection-characterized activity: during this activity, actors have to make decision. It leads to at least one alternative. - Execution-characterized activity: the carrying out of this activity leads impact on the operative sub-system. It corresponds to the switch from intention to act, involving some resources of the enterprise. 6.

A case study

We illustrate the common concept of activity presented above with an example extracted from previous works led by our research team [2]. The studied enterprise is specialised in the design and production of industrial transformers using copper wires. To stay competitive, the enterprise must optimise production delays, costs and product quality. Consequently, the production manager must adapt its design processes according to the copper price which may vary frequently. The design engineer must choose between two predefined design processes which can be modified in real time: if the copper price is low, the process consists in using thick-diameter wire of copper and a lower number of twists (design process A), if the copper price is high, the transformer is manufactured with copper wires of thin diameter and a more important number of twists (design process B). The representation of the different activities is described in the figure 4 as a process considering the copper price with a sub-level decomposition of the rescheduling activity. [fig 4] in the appendix illustrates the different activities involved in that process. The different actors concerned, their role in the decision-making process, the performance indicators and skills of each activity are stressed. 7.

Perspectives

In this paper, we have presented how the concept of activity could unify the control modelling. In order to fulfil this objective, our ongoing researches consist in the development of these two successive steps : -

-

Firstly, refining the characterisation of the notion of activity. This step consists in the study of the decision processes made by actors of design and production control. Indeed, it is necessary to take into account the occurrence of disturbances (like the change of functional specifications) in order to characterise the triggering of activities; Secondly, integrating activity into an approach for modelling collective decision processes in production and design control. Such an objective requires to identify the set of actors involved, their

interactions, and the interrelations between the activities of these domains. From those points of view, we intend to rely on the researches which process the notions of (i) traceability of the decision processes, to benefit from future reusing, and (ii) skills to describe the knowledge of the actors in order to represent their skills and thus resulting in a better allocation of the human resources. In order to validate our research, we work on the implementation of the common concept of activity in an example, which takes into account problems of communication and reactivity in decision processes control. Thus, we project to apply our modelling approach on a scenario in concurrent engineering context. Our aim is to control and analyse the processes of the decision making, developed to face disturbances triggered by a design issue, in a manufacturing process. Finally, our purpose is to provide a common language for actors from different culture by developing a common modelling approach for the two domains. This language will be built on a common concept of activity, the tools and concepts evolved to analyse decision processes in (i) the production systems control framework presented in [2], (ii) the concurrent design developed by [3]. 8.

References

[1] Projet pluri-équipes, "Intégration des activités nonstructurées dans la modélisation des systèmes de productionACNOS, rapport final ", Février 1997. [2] AYADI.K, CAUVIN.A, KIEFFER.J.P, "Towards an approach of structuring control systems for production management", systems analysis modelling and simulation revue, N°324, Vol.32, pp 257-283, 1998. [3] BOUCHER.X, KIEFFER.J.P, "Gestion des compétences liée à la mise en œuvre de l'ingenierie concourante", Actes du 3ème C.I.G.I, Montréal, Canada, 26-28 Mai, 1999. [4] GIRARD.P, EYNARD.B, DOUMEINGTS.G, "Proposal to control the systems design process: Application to manufactured products", Proceedings of the 2nd international conference on integrated design and manufacturing in mechanical engineering, UTC, Compiegne, France, 27-29 Mai, 1998. [5] MAYER.R.J, MENZEL.C.P, PAINTER.M.K, de WITTE.P.S, BLINN.T, , PERAKATH.B, "Information integration for concurrent engineering, IICE, IDEF3 Process description capture method report", college station (Texas) : Knowledge based systems, september, 1995, 224p. [6] MEGARTSI.R, AYADI.K, CAUVIN.A, KIEFFER.J.P, "L’activité : vers un concept fédérateur pour l’analyse de la conduite des systèmes de production et des processus de conception de produits", Actes du 3ème Congrès International de Génie Industriel, Montréal, Canada, 26-28 Mai, 1999. [7] PRASAD.B, "Concurrent Engineering Fundamentals", Volume II : Integrated Product and Process Organisation", New Jersey : Prentice Hall PTR, 1997.

9. Appendix Performance Indicators(PI) required cost x quality x x time

effective x x x

PI following scheduling

identifying price of the copper

I

X

1.1

E 4.1

Tech-manager

PI

purchase-manager

Skill Available human resource

Finding supplier Code

E

Skill 2.1 Purchase manager

O

Required skill of human resource

PI

Skill

X PI

Re-scheduling

E 5.1

Tech-manager

Fixing a stock

E Skill

3.1 Purchase manager

Skill

PI Applying design process A

E 5.5.1

Design manager

PI

PI

Identifying products

I

Consulting position in the scheduling

5.1.1 Production manager

PI I

Ending the work in progress

5.2.1 Project manager

PI E

Changing production S process

5.3.1 Tech-manager

Skill

5.4.1

X

Production manager

PI Applying design process B

E 5.6.1

Skill

Skill

Skill

Skill

Design manager

Skill

Figure 4: Representation of the process "taking into account the copper price".