Curriculum vitae

Sali Mustapha

Maître de conférences
DRM

mustapha.saliping@dauphinepong.fr
Tel : 0144054033
Bureau : A720

Publications

Articles

Sali M., Sahin E. (2016), Line feeding optimization for Just in Time assembly lines: An application to the automotive industry, International Journal of Production Economics, 174, p. 54–67

The performance of mixed-model assembly lines used in sectors such as the automotive industry depends on the availability of a large number of components that have to be supplied to the line on time and at minimum cost. In such settings, components may have different features such as volume, weight, bill of material coefficient, etc. Additionally, a given component may have several alternative variants among which a single one is used in the assembly of end products. Each variant is thus characterized by a varying degree of usage rate. Hence, the diversity of parts requires the selection of the best assembly line feeding mode that aims at minimizing the average total operating cost which mainly consists of labor costs associated with parts preparation before assembly, transportation to the line, picking operations during assembly, as well as parts storage cost. This paper proposes an optimization model that assigns each individual component to the most efficient line feeding mode among three alternatives which are line stocking, kitting and sequencing modes. The developed mixed integer program is applied to a first tier supplier plant in the automotive sector. Based on this model, insight is gained on the trade-off to be considered when deciding the more appropriate line feeding mode for each individual component and how system parameters impact this trade-off.

Chatras C., Giard V., Sali M. (2016), Mass customisation impact on bill of materials structure and master production schedule development, International Journal of Production Research, 54, 18, p. 5634-5650

The customisation of a vehicle on a production line results from the assembly of several hundreds of alternative components. Bills of Material (BOM) are usually made for operational purposes (to define the list of components to be assembled on a vehicle), for planning (to anticipate component procurement needs) and for commercial needs (guide customer choice and prepare sales forecasts). In the automotive industry, the diversity of end products which results from this combinatorial process (several millions) is such as no solution proposed in the literature allows to easily list the BOMs for all these vehicles. In our paper, we describe and theorise a solution used for many years by several carmakers, which consists in introducing a commercial description of the products in addition to their common organic representation. We show that this solution is an extension of the generic and modular BOM theory, recognised to be the most advanced solutions proposed in the literature. While fully meeting commercial and production control needs, this product description paradigm has limitations when it comes to determining the Master Production Schedules (MPS) at end product level beyond the frozen horizon, as performed by several automakers. An alternative approach defining MPSs at alternative component level is proposed in the literature. However, this approach also has several drawbacks that we point out. To overcome these issues, we propose a number of possible strategic and organisational tracks of improvement.

Chatras C., Giard V., Sali M. (2016), Impacts d'une forte diversité sur la construction des nomenclatures et des plans directeurs de production, Journal européen des systèmes automatisés, 49, 1, p. 55-91

En production de masse fortement diversifiée, le très grand nombre de produits finis et les contraintes techniques et commerciales, limitant les combinaisons des composants alternatifs (CA) assurant la diversité, rendent très complexe la création, la maintenance et l'utilisation de l'ensemble des références de nomenclature associées à chacun des produits finis. Cette complexité explique que le client ne puisse définir sa demande à partir de CA et la difficulté d'établissement de prévisions comme rciales au niveau des produits finis ou des CA. Depuis une vingtaine d'années, certains constructeurs automobiles utilisent une description originale des produits finis, qui ne semble pas avoir été théorisée. Elle décrit un produit fini par un ensemble de prestations alternatives, ce qui correspond à une vision fonctionnelle. Le passage à une description organique (CA) - incontournable pour l'établissement du plan directeur de production (PDP) et la production - s'obtient par le biais de prédicats. Cette approche est décrite et illustré e à partir d'un cas réel emprunté au secteur automobile. Elle répond aux besoins commerciaux et opérationnels mais pose de redoutables problèmes dans l'établissement du PDP au-delà de l'horizon gelé.

The creation, updating and pointing of the full set of the Bills of Materials corresponding to each end-product is extremel y complex in a mass customization context, because of the great variety of both end products and components and the existence of commercial and technical constraints between th e alternative components (AC) ensuring that variety. Customers can no longer define their product requirements by specifying a list of components and sales department cannot issue fo recasts either at end-product or component level. Within the last two decades, to address these difficulties, some carmakers developed a new approach to describe products. This product description, which doe s not seem to have ever been theorized in the literature, is based on product market features through the concept of alternative services. The conversion to a physical Bill of Materials - that remain mandatory for the Master Production Schedule definition and for the production activities - is done through the use of predic ates. In this paper we first develop this new approach and then describe an application for drawing up MP Ss, that is not straightforward beyond the frozen horizon.

Sali M., Giard V. (2015), Optimal stock-out risk when demand is driven by several mixed-model assembly lines in the presence of emergency supply, International Journal of Production Research, 53, 11, p. 3448-3461

This article focuses on the calculation of the optimal stock-out risk when resorting to emergency supplies is allowed. The studied context is a supply chain where several distant mixed-model assembly lines trigger the demand of a component used for assembly. When the cumulated lead time of a component exceeds the frozen horizon of the material requirement planning (MRP) of at least one of the assembly lines, its demand becomes stochastic. This leads to a mix between production to stock and production to order for the component of interest. The periodic replenishment policy is designed to address demand uncertainty in consistence with MRP information. To prevent stock-out propagation along the supply chain, the emergency supply is used as a last resort. The calculation of the optimal order-up-to-level and its associated optimal stock-out risk are based on a single-period trade-off between holding cost and emergency supply cost (transportation and production).

Sali M., Giard V. (2015), Monitoring the production of a supply chain with a revisited MRP approach, Production Planning & Control, 26, 10, p. 769-785

This paper examines the centralised monitoring of an upstream supply chain (USC) in the context of a mass customisation production system. We propose an adaptation of the material requirement planning framework to manage demand uncertainty at each stage of the USC. New analytical relations that exploit structural information on demand beyond the frozen horizon are developed for efficient decision-making and appropriate information sharing among the productive supply chain units. A numerical example to illustrate the implementation of the new analytical relations is provided, and a synthesis of benchmarks performed in the automotive industry is presented.

Sali M., Sahin E., Patchong A. (2015), An empirical assessment of the performances of three line feeding modes used in the automotive sector: line stocking vs. kitting vs. sequencing, International Journal of Production Research, 53, 5, p. 1439-1459

In high diversity, Just In Time production environments, mixed-model assembly lines aim at producing a large amount of end products that use alternative variants of components used in the assembly process. Therefore, the availability of parts required at the border of the assembly line is critical for the production process. To ensure the availability of parts needed on the line, three line feeding modes are commonly used in practice: line stocking, kitting and sequencing. The relative performance of each mode, in terms of average total operating cost, depends on various factors such as the diversity of components, parts' physical features (i.e. weight and volume), to name a few. The impact of such parameters on the performance of line feeding modes remains still weakly studied in the literature. Through an empirical approach based on a real case, our study aims at evaluating the average total cost pertaining to each mode. The total cost includes parts preparation before assembly, picking, in-plant transportation and storage costs. A thorough comparative study enables then to characterise situations which make one mode less costly compared with the others. Each situation corresponds to specific values taken by parameters that are considered in the cost formulation.

Giard V., Sali M. (2014), Optimal stock-out risk for a component in the case of emergency supplies, Journal Européen des Systèmes Automatisés, 48, 1-3, p. 55-77

On s'intéresse ici à la détermination de la probabilité de rupture optimale d'un composant utilisé par un ou plusieurs modules alternatifs montés sur une ou plusieurs lignes d'assemblage. Le contexte étudié est celui d'une chaîne logistique dédiée à une production de masse de produits fortement diversifiés, comme celui rencontré dans l'industrie automobile. L'approche de la MRP (Material Requirement Planning) est adaptée pour piloter cette chaîne dans laquelle l'éloignement des unités de production conduit à un panachage de production à la commande et de production pour stock du composant auquel on s'intéresse. Pour éviter toute propagation des conséquences d'une rupture de stock en aval de la chaîne logistique, un approvisionnement d'urgence est toujours déclenché avant son occurrence. La définition du stock de sécurité - et du risque de rupture associé - se base sur un modèle monopériode intégrant le coût des composants inutilisés et les différents coûts induits par l'approvisionnement d'urgence (transport et production). Les solutions analytiques proposées sont illustrées numériquement et une analyse des conditions de dominance économique des deux principales politiques alternatives d'approvisionnement d'urgence est menée.

This article focuses on the calculation of the optimal stock-out risk for a component, which is used by alternative modules mounted on several assembly lines. The studied context is a supply chain dedicated to the mass production of highly diversified products, which is common in the automotive industry. The Material Requirement Planning (MRP) approach is adapted for the monitoring of this chain; however, the distance between the production units leads to mix between production to stock and production to order for the component of interest. To prevent stock-out propagation along the downstream part of the supply chain, use of an emergency supply is triggered prior to its occurrence. The definition of the optimal safety stock and the associated optimal stock-out risk, are based on a monoperiod model that considers the cost of a safety stock and the costs incurred by the emergency supply (transportation and production). The analytical solutions are illustrated and two alternatives emergency policies are compared.

Sali M., Giard V. (2013), The bullwhip effect in supply chains: a study of contingent and incomplete literature, International Journal of Production Research (IJPR), 51, 13, p. 3880-3893

The bullwhip effect (BWE) describes a phenomenon that involves the increasing amplification of demand variability along a supply chain (SC). The BWE has been a subject that has received continuous attention from researchers over the past 15 years and is a concern for SC managers because it is a major cause of efficiency and effectiveness loss in SCs. Information sharing between actors in an SC is usually considered to be one of the primary means to minimise the BWE. Approximately 50 articles published in major journals on these topics are studied in this article. An analytical framework is used to highlight the contingent character of the conclusions proposed by the authors. In this review, we identify the existence of significant gaps in the literature, especially concerning the BWE when it occurs in the productive part of the SC.

Giard V., Sali M. (2012), L'effet coup de fouet dans la chaîne logistique : une littérature contingente et incomplète, Revue française de gestion industrielle, 31, 2, p. 24-42

L'effet coût de fouet (ECF) correspond au phénomène d'amplification croissante de la variabilité de la demande en remontant le long de la chaîne logistique (CL). Il fait l'objet d'une attention soutenue par les chercheurs depuis une quinzaine d'années et préoccupe les gestionnaires de CL parce qu'il est une cause importante de perte d'efficacité et d'efficience dans la CL. Le partage d'informations entre acteurs de la CL est généralement considéré comme l'un des principaux moyens de réduction de l'ECF. Une cinquantaine d'articles portant sur ces thèmes et parus dans les revues considérées comme majeures ont été étudiés. Les grilles d'analyse utilisées permettent de mettre en évidence le caractère très contingent de nombreuses conclusions et l'existence de lacunes importantes dans les investigations conduites, notamment en ce qui concerne l'ECF se produisant dans la partie productive de la CL.

Giard V., Sali M. (2012), Pilotage d'une chaîne logistique par une approche de type MRP dans un environnement partiellement aléatoire, Journal européen des systèmes automatisés, 46, 1, p. 73-102

On s'intéresse au pilotage par une approche MRP, d'une chaîne logistique dédiée à laproduction de masse de produits fortement diversifiés. Il s'agit plus précisément de coordonner laproduction d'un ensemble de modules alternatifs montés sur un poste d'une ou plusieurs lignesd'assemblage final ainsi que celle des composants que ces modules comportent. Les niveaux deproduction de ces lignes sont stables. Les caractéristiques structurelles des demandes finales sontstables ou prévisibles. La production d'un composant est tirée par des réquisitions du Plan Directeurde Production (PDP) hors horizon gelé ne peut pas se faire à la commande dans le cadre classiquede la MRP. On propose des solutions analytiques permettant de piloter la partie productive de lachaîne logistique par une approche MRP autorisant des productions partiellement ou totalementpour stock. On introduit également d'autres relations permettant de prendre en compte les aléas dequalité en production. Toutes ces solutions sont illustrées numériquement.

This article focuses on the monitoring of a supply chain dedicated to the massproduction of strongly diversified products. In particular we are interested in the part of thischain that contributes to the production of a set of alternative modules assembled on a workstation of one or several assembly lines, whose production levels are stable. The customers'demand structures are assumed to be known. When the production of a component is drawn byrequirements of the MPS that are beyond the frozen horizon, this component cannot be built-toorder.We provide and illustrate analytical solutions, adapted from the periodic order policies,to define scheduled orders of components to be produced partly or wholly to stock. Additionalrelations are provided to face disturbances induced by quality problems.

Chapitres d'ouvrage

Sali M. (2015), Le lean dans l'industrie manufacturière : évolutions et nouveaux enjeux, in Dauphine Recherches en Management (DRM) ., Nogatchewsky G., Perret V. (dir.), L'état des entreprises 2015, Paris?, La Découverte, p. 87-102

Communications

Chatras C., Giard V., Sali M. (2015), High variety impacts on Master Production Schedule: a case study from the automotive industry, INCOM 2015, 15ième IFAC/IEEE/IFIP/IFORS Symposium, Ottawa, Canada

In mass customization production systems, end-products diversity proceeds from the combination of hundreds of alternative components (ACs) involved in the final assembly stage. Since diversity is very important, forecasts cannot be done directly at the end-product level for production planning purposes. Thus, it is generally agreed in a mass customization context that Master Production Schedules (MPSs) have to be defined for each AC rather than each end-product. Nevertheless, some issues may hamper this approach. First, the MPSs have to be consistent with the technical and commercial constraints that prevent some ACs combinations. Second, due to longer lead times resulting from globalization, production planning must accommodate with uncertainty over the frozen horizon. To address these issues in the automotive sector, a new way to represent product through the concept of the Alternative Services (ASs), which refers to the alternative commercial features offered to the customer, has been introduced by some carmakers. This paper proposes a description of this approach and discusses its relevance regarding the issues of the classical production planning process. Some suggestions are proposed to overcome the limitations of the approach based on the ASs.

Chatras C., Giard V., Sali M. (2015), High variety impacts on Bill of Materials Structure: Carmakers case study, 2015 IFAC Symposium on Information Control in Manufacturing (INCOM 2015), Ottawa, Canada

The great variety of offerings from companies engaged in mass customization such as in the automotive sector, implies millions of end products. As in the case of carmakers, this variety proceeds from the combination of dozens of optional or alternative components. A number of technical or commercial constraints prevent most combinations of alternative components (ACs). In these conditions, the creation and updating of the full set of the Bills of Materials corresponding to each end-product is extremely complex. Additionally, customers may not define their product requirements by specifying a list of alternative components because of the number of alternative components (including some entirely unknown to them) and their interdependence. In the past 25 years, scant research has been dedicated to Bills of Materials representation problems in this context. The most exciting ones use predicates to formalize combinatorial constraints between ACs and introduce a 'generic Bill of Materials' concept to avoid an exhaustive description of all end products. Additionally, product description for customer purposes was streamlined through a further concept - the set of alternative services (SAS). This approach, implemented by several carmakers in the last fifteen years, superseded the traditional approach to bills of materials, though it never gave rise to the theoretical exercise described here that involves an analysis of its impact on bills of materials and on operational and tactical decisions. It describes the product based on its market features, before moving on to the actual ACs combinations based on predicates for alternative services (ASs). We show how some carmakers successfully use the AS concept to support sales at configurator level and address operational production issues.

Sali M., Giard V. (2014), Optimal stock-out risk for a component mounted on several assembly lines in case of emergency supplies, 7th IFAC Conference on Manufacturing Modelling, Management, and Control, 2013, Saint-Petersbourg, Russie

This article focuses on the calculation of the optimal stock-out risk for a component, which is used by alternative modules mounted on several assembly lines. The studied context is a supply chain dedicated to the mass production of highly diversified products, which is common in the automotive industry. The Material Requirement Planning (MRP) approach is adapted for the monitoring of this chain; however, the distance between the production units leads to mix between production to stock and production to order for the component of interest. To prevent stock-out propagation along the downstream part of the supply chain, use of an emergency supply is triggered prior to its occurrence. The definition of the optimal safety stock and the associated optimal stock-out risk, are based on a monoperiod model that considers the cost of a safety stock and the costs incurred by the emergency supply (transportation and production). The analytical solutions that are dependent on these costs are illustrated in this study.

Sali M., Giard V. (2012), Monitoring of the upstream part of a supply chain dedicated to customized mass production with a revisited version of MRP, ILS 2012, Quebec, Canada

This article focuses on the monitoring of a supply chain dedicated to the mass production of strongly diversi-fied products. Specifically, we are interested in the part of this chain that contributes to the production of a set of alter-native modules assembled on a work station of one or several assembly lines, whose production levels are stable. The MRP approach is adopted for the monitoring of this chain. The distance between the production units leads to a mix between production to stock and production to order. In this article, we establish the relations that allow us to define, in a steady state, the quantities to produce that address the requirements of the Master Production Schedule and that are partially or completely random to limit the stockout risk to a very low predetermined level. We will distinguish two cases by accounting for, or not accounting for, problems that are related to quality.

Documents de travail

Sali M., Giard V. (2011), Production à la commande et production pour stock dans un environnement MRP, Cahiers du LAMSADE, Paris, Université Paris-Dauphine, 9

Cet article s'intéresse au pilotage d'une chaîne logistique dédiée à une production de masse de produits fortement diversifiés et plus particulièrement à la partie de cette chaîne qui contribue à la production d'un ensemble de modules alternatifs montés sur un poste de la ligne d'assemblage final. L'approche de la MRP est retenue dans le pilotage de la chaîne mais l'éloignement des unités de production oblige à un panachage entre production pour stock et production à la commande. La littérature disponible sur les stocks de sécurité n'est pas d'un grand secours opérationnel. On examinera la relation qui existe entre horizon gelé et le point de pénétration de commande et on établira les relations analytiques qui lient les lancements en production d'une référence et les réquisitions du PDP qui les utilisent. On établira les relations qui permettent de définir, en régime de croisière, les quantités à lancer pour satisfaire une demande partiellement ou complètement aléatoire, avec ou sans problème de qualité, afin de limiter le risque de rupture de stock à un niveau prédéterminé.

This article focuses on the monitoring of a supply chain dedicated to a mass production of strongly diversified products and more particularly in the part of this chain which contributes to the production of a set of alternate modules assembled on a work station of the assembly line. The approach of the MRP is adopted in the monitoring of that chain but the distance of the production units leads to a mixing between production to stock and production to order. The literature available on safety stock does not help in decision making. The relation that exists between frozen horizon and Order Penetration Point in MRP will be examine and the analytical relations which links planned orders of a reference to the requirements of the MPS which uses them, is established. Then we set the relations which make it possible to define, in the steady state, the quantities to produce to cope with requirements that are partially or completely random, taking into account or not the problem of quality, in order to limit the stockout risk to a predetermined level.

Retour à la liste