81 Seiten, Note: 1,0
List of Figures
List of Tables
List of Abbreviations
List of Symbols
2 Automotive planning processes
2.1 Product development process
2.1.1 Description of a product development process
2.1.2 Importance of the product development process
2.1.3 Planning horizons
2.2 Reference planning landscape for automotive OEMs
2.3 Role of platform strategies in automotive planning
2.3.1 Platform definitions and example
2.3.2 Benefits of platform strategies
2.3.3 Platform decisions and implications
2.4 Strategic operations planning tasks
3 Integration of the operations perspective into platform planning
3.1 Integration challenges and requirements
3.2 Literature review
3.3 Research gap
4 Reference MILP model
4.1 Problem setting
4.2 Objective function
5 Numerical study
5.1 Data set
5.2 Single-stage optimization
5.3 Two-stage incremental optimization
6.2 Outlook and recommendations
Figure 1: Reference planning landscape for automotive OEMs (Jana and Grunow, 2016)
Figure 2: Comparison of cost drivers
Figure 3: Comparison of number of platforms
Figure 4: Comparison of production sites per platform
Table 1: Overview of modeling approaches dealing with platform planning
Table 2: Indices of the numerical study
Table 3: Allocation of vehicle derivatives to platform concepts in a single-staged approach
Table 4: Allocation of vehicle derivatives to platform concepts in a two-staged approach
Abbildung in dieser Leseprobe nicht enthalten
In order to react on a broader product portfolio and shortened product life cycles, the usage of platform concepts gains an increasing importance for automotive OEMs to establish cost and time savings. However, current industry best practices and existing literature do not integrate the operations perspective into the planning of automotive platforms to a sufficient extent. Therefore, this thesis discusses the role of the strategic operations planning for platform planning. This results in a mixed-integer linear programming approach, integrating the operations perspective into the planning of automotive vehicle platforms. The overall objective is to allocate existing vehicles to an existing set of platform concepts, while minimizing associated cost. Based on a numerical case study the use of the newly established model is compared with an incremental platform planning approach. Finally, the benefit of the model is evaluated before the integration of further potential operations aspects gets discussed.
Today, automotive original equipment manufacturers (OEMs) experience an increasing complexity of their planning and development activities due to shorter life cycles and broader product portfolios (Pil and Holweg, 2004). In order to minimize the additional costs for this diversity, vehicle platforms become more important to realize saving potentials from synergy and scale effects in development, production and sourcing across several car models (Jana and Grunow, 2016). A product platform in general can be considered a certain set of vehicle components, which are used identically across several products. This results in cheaper, more flexible and more responsive design, planning and manufacturing processes (ElMaraghy et al., 2013). Therefore, OEMs increasingly implement so-called multi-platform strategies, establishing several platforms, e.g. front wheel drive platform, rear wheel drive platform, SUV platform, etc. (Jana and Grunow, 2016). Every single platform is used for a distinct set of vehicles, the so-called platform derivatives, e.g. sedan, sports coupé, station wagon, etc.
However, the use of platform technologies may, among others, decrease the degree of diversity for a certain vehicle and therewith result in lost sales and further disadvantages.
Unlike the individual car development process, the platform development is usually reviewed on a cyclic base every year with a planning horizon of more than ten years (Jana and Grunow, 2016). The car development process itself is separated into several stages, merging into a long-term design project. Even though operated independently, the processes of platform planning and individual car development are strongly connected. In order to avoid a major loss of synergy effects, additional investment costs or lost sales due to late changes to a specific platform, these processes have to be aligned. Besides the consideration of the car’s design and the target costing in the platform planning, this mainly comprises the operations perspective of the production and supplier network planning and the manufacturing technology design on a strategic level.
However, thus far both research and practice have mostly focused on effects on product design and neglected the implications on the OEM’s operations when planning multi-platform strategies. Even though the need for the holistic integration has been understood by most OEMs, the current industry practices still offer room for improvement. Therefore, this thesis will focus on the integration of the operations perspective into automotive platform planning.
The purpose of this thesis is to analyze the current research efforts on the integration of the operations perspective into automotive vehicle platform planning and provide an optimization model for future platform planning. In particular, the objectives of the thesis are:
- to provide a basic understanding of automotive planning processes in general and the relevance of platform concepts.
- to identify the current research gap.
- to develop a mixed-integer linear program (MILP) based model, integrating operations factors into automotive platform planning.
- to proof the benefit of this new model with a sample data set in comparison to an incremental planning approach.
- to point out further room for improvement and suggest further measures.
In order to understand automotive planning processes, firstly the importance of the product development process in combination with its integrated operations tasks will be explained in Section 2.1. Based on a provided reference planning landscape for overall automotive planning tasks in Section 2.2, the platform planning and strategic operations tasks are discussed in detail in Section 2.3 and Section 2.4. Chapter 3 provides an overview about the identified challenges and requirements for integrating the operations perspective into platform planning and about the existing literature. Consequentially, the current research gap is identified and described in Section 3.3. Following on that, the corresponding operations factors for an operations research modelling approach are documented in the introduction of Chapter 4, which leads to the MILP-based model itself, demonstrating the vehicle-to-platform allocation and cost impacts from Section 4.1 to Section 4.3. After the establishment of this theoretic and generic approach, the application of this model with a numerical example is shown and discussed in Chapter 5. Based on a comparison of the application of a single-stage model to an incremental approach, the value of the integration of the operations perspective into the planning of automotive vehicle platforms is proven in Section 5.4. Following up on these results, Chapter 6 consists of a summary in Section 6.1 and an outlook in Section 6.2 in order to highlight certain improvement potentials for the created model as well as to suggest further measures.
A process in general is defined as a sequence of activities in a logical connection with a distinct start and end date (Schönmann, 2011, p.79). It requires to transform a certain input into an output through a value adding activity. The input, value adding and the output have to be measurable. Concerning the automotive industry Wildemann (2004, p. 266) proposed a reference model for a car development process consisting of six different stages. For this industry, the input is represented in form of a certain customer need for a new vehicle and the output represents this new vehicle. Wildemann’s process approach consists of an idea generation stage, a product definition stage, a concept development stage, a product development stage, a start-up stage and a series production stage. Within the idea generation stage ideas and potential innovations are identified and evaluated based on market research studies. The ideas are evaluated concerning the factors of technical feasibility and economical aspects. Furthermore, potential requirements and objectives about the new vehicle’s characteristics are defined. Then, the product definition stage results in a project development request, containing the customers’ requirements for the new product concerning technical aspects as well as pricing issues. Moreover, a first make-or-buy analysis and a pre-selection of potential production locations are reported. Next, the concept development stage defines the various individual functional product units, from which some are merged into modules or systems. In addition, a detailed analysis of the procurement market, resulting in precise make-or-buy decisions, is conducted. This includes a supplier pre-selection and initial negotiations. Following on that, within the product development stage, the actual product is constructed and evaluated as a prototype. At the same time, the manufacturing and logistic processes are planned. Besides, contracts with suppliers are build. Then, the start-up stage provides an approval for the developed product in order to start the mass production, which is finally operated in the series production stage. Besides, the series production stage covers additional optimization measures for the product itself and the introduced manufacturing and processing techniques.
The product development stages have a high impact on the effectiveness of development and efficiency of resources, especially at an early stage. As the automotive OEMs have to face high competition on highly complex products, technical development in form of innovations is crucial (Schönmann, 2011, pp.79-81). These innovations have to fit the customers’ expectations. Therefore, the product development process has to be market orientated. Furthermore, short lead times and fast development are important because demand anticipation gets more complex with an increased planning horizon. Long-lasting demand planning horizons would lead to an inaccurate implementation of customer requirements. Concerning the company internal perspective, it is necessary to determine the product, production, logistics and disposal costs for the new vehicle and associated operations within the development process in order to enable reliable financial planning. In addition, quality characteristics are a fundamental planning factor. The elimination of defects within the development process roughly costs ten percent of what the elimination would cost after the market launch. Hence, it is essential to define and implement quality measures as early as possible. Overall, within the product development process, a profitable balance of cost, quality and timing is important. From all these mentioned factors follows that the car development process represents a shareholder value with enormous importance because of its high influence on costs and revenues.
The planning landscape is divided into three planning horizons: the strategic long-term planning, the tactical mid-term-planning and the short-term and operational planning (Fleischmann et al., 2015). Firstly, the strategic long-term planning defines the design and structure of the supply chain with a planning horizon of several years. The included planning tasks are mainly about long-term supplier relations, the plant locations and production systems, the physical distribution structure and a strategic sales plan. Secondly, the tactical mid-term-planning determines the time and quantity of the flow of goods and use of resources in a given supply chain on an approximate level for the next six to twenty-four months. The relevant tasks cover a material requirement planning, personnel capacity planning, production scheduling and capacity planning, the distribution planning between warehouses and stock planning, as well as the mid-term sales planning on a region level. Thirdly, the short-term and operational planning comprises the final operations task on a high detail level. This includes, amongst others, the short-term sales planning, warehouse replenishment and transport planning, lot-sizing and machine scheduling as well as short-term personnel planning. The planning period for the short-term and operational planning spans from a few days up to three months.
In order to display the overall planning processes for the development of new products at an automotive OEM, it is necessary to integrate the discussed product development process into a superordinate planning landscape. In general, the landscape differentiates between the two different planning processes of design projects and cyclic planning (Jana and Grunow, 2016). Design projects are characterized as long-term projects with a dedicated team and a distinct end date. They address the organization and implementation of the product design to develop new vehicles, the process technology design to develop new manufacturing technologies and the network design and ramp-up planning in order to adapt the production and sourcing network for a new vehicle. On the other hand, cyclic planning is operated in defined cycles. Firstly, it consists of the product planning to review the current product portfolio and strategy in order to propose changes to react on the market development. The second element, the demand planning, provides a global long-term demand forecast on an annual basis (Clark and Fujimoto, 1991, pp. 97-128). Thirdly and finally, the tactical sales and operations planning (S&OP) aligns supply and demand. In addition, it coordinates the operational network and production planning with the business strategy (Tuomikangas and Kaipia, 2014). The operational order fulfillment can also be considered a cyclic planning process.
Jana and Grunow (2016) provide a detail reference planning landscape for automotive OEMs, integrating the product development process into a planning landscape of design projects and cyclic planning, which is displayed in Figure 1. The horizontal axis determines the time, when and for how long to execute the included planning tasks. On the vertical axis the planning tasks are sorted by design project and cyclic planning.
Similar to the product development process approach by Wildemann (2004), presented in Section 2.1, the considered reference process of Jana and Grunow (2016) splits the actual car development processes into concept design, specification and product prototyping. In concept design, research and development (R&D) defines the new vehicle concerning the aspects of technology, working principles, form, size of product and further specific requirements (Ulrich and Eppinger, 2008). Once the definition of these information is finished and described in a concept book, the vehicle specification phase starts. During that processing stage, all single components of the vehicle are technically described in detail (Jana and Grunow, 2016). The development process finally ends with the product prototyping eighteen months before the SOP.
Along with the start of concept design, the task component allocation determines a roadmap for make or buy decisions for modules and components based on the expected life cycle volume, costs and preliminary vehicle concept, and already states a set of potential suppliers. After a certain component is fully specified, the final make or buy decision can be made and in addition suppliers can be evaluated and finally selected. The responsible buyers can either create finalized contracts in order to already define the volumes for the entire life cycle or design annual contracts later on.
Besides, the process of target costing defines the financial boundaries for the product development approximately five years before SOP (Jana and Grunow, 2016). Roughly at the same time, location planning and strategic ramp-up planning start to evaluate and propose potential production sites for the vehicles and their platforms. While the location planning deals with the platform-to-plant allocation, the strategic ramp-up planning determines the vehicle-to-plant allocation and defines the platform life cycle plan. In parallel, process technology design defines the boundaries for the future manufacturing costs and feasibility by developing new process technologies.
Abbildung in dieser Leseprobe nicht enthalten
Figure 1: Reference planning landscape for automotive OEMs (Jana and Grunow, 2016)
Based on the location planning, the task facility design is concerned with the plant layout planning for both existing and new facilities. At the same time, capacity planning covers the detailed planning of the production processes at each production stage, namely press shop, body shop, paint shop and final assembly, for all relevant sites (Volling et al., 2013). Furthermore, the capacity planning covers the selection of equipment suppliers and the decision about capacity investments. The completion of these tasks initiate the construction of a prototype plant and the plant certification process (Jana and Grunow, 2016) ending in the final plant testing. In parallel, around six to twelve months before the SOP, the tactical ramp-up planning defines the detailed production ramp-up curves for the new vehicle and the market launch scheme.
For the cyclic planning, the already mentioned product planning can be divided into platform planning and module planning on a strategic level. Platform planning defines and reviews the future vehicle portfolio with a planning horizon of more than ten years and is reviewed every year. In module planning the generic module structure across all vehicles, the type of each module and the module life cycle are defined. The review of the existing module strategy is deducted in cycles of six months to one year. Even though platform and module strategies are related to each other, they have to be managed separately (Kauder, 2008).
The strategic volume planning determines the central production forecast on a yearly basis with a planning horizon of more than ten years (Jana and Grunow, 2016). The planning is operated by the sales and R&D departments. It lasts several months and is fixed two years before the start of production (SOP) for every vehicle.
The cyclic annual and monthly demand planning in the automotive industry combine to the S&OP. The annual S&OP cycle comprises the annual demand planning, the budget planning (BP), the annual allocation planning (AP) and annual material requirement planning (MRP). Based on the demand planning, BP involves representatives from the central sales, production and procurement departments and is considered the central S&OP task (Meyr, 2004). Monetary budgets and central annual production plans are derived from BP (Jana and Grunow, 2016). Furthermore, it influences the annual MRP, which determines the supply plan for the next year.
The monthly S&OP cycle comprises the monthly demand planning, the master production planning (MPP), the monthly AP and the monthly MRP. The monthly MPP represents a more accurate version of the annual BP with a planning horizon of three months to one year (Meyr, 2004). It is mainly concerned with the matching of supply and demand across the entire network (Jana and Grunow, 2016). Furthermore, it provides the central monthly production and sales plans and redefines the workforce schedule for the next months. Based on the MPP the monthly MRP provides the detailed component supply plan for the next months. The planning horizon is between three and 12 months (Staeblein and Aoki, 2015).
The remaining displayed four boxes of short-term MRP, master production scheduling (MPS), sequencing (S) and order promising (OP) conflate to the order fulfillment process, operated on a weekly to daily basis. The MPS, based on the monthly MPP, integrates the actual orders into the production plans. Within OP, orders are assigned to a production site and integrated into the site’s master production schedule (Volling and Spengler, 2011). Based on that, S finally determines the exact car sequence for each line using a specific sequencing tool (Holweg et al., 2005).
The various definitions for platforms in the existing literature are not consistent, but they mostly apply to the automotive industry (Mufatto and Roveda, 2000). Therefore, in this thesis, platforms are considered a shared set of components common to a number of different vehicles (Weber, 2009, pp.4-6), in order to create a consistent understanding. The platform architecture or structure can either follow a module-based or a scale-based platform design approach (Bhandare and Allada, 2009). A module-based platform consists of several modules, consisting of several components and parts. The platform can be modified by adding, removing or substituting one or more functional modules from or to the platform (Simpson, 2004). For a scale-based platform design different design variables of the platform are varied in order to create diversified platforms in form of variants (Bhandare and Allada, 2009). These variants consist of common and modified, so-called scalable, design variables.
One of the most famous platform examples is the Golf-platform, developed and utilized by the Volkswagen Group (Weber, 2009, pp.4-6). It is shared by four different brands on thirteen different vehicle models. It consists of parts of the powertrain, the steering, the suspension, the lower body and the interior trim. Differentiation is ensured by varying exterior bodies and interior trims, though some parts only differentiate by the attached brand label.
Three commonly used strategies for product platforms are based on a market segmentation grid by Meyer (1997) and Meyer and Lehnerd (1997). In a market segmentation grid the areas of the total market can be differentiated according to the two different dimension of market areas and performance values. While the strategy of horizontal leveraging penetrates several market areas on the same performance level with the same platform, the strategy of vertical leveraging is applied with one platform within one market area used across a range of performance values and customer requirements (Seepersad et al., 2000). Hence, for the horizontal leveraging the platforms vary across the performance levels and for the vertical leveraging the platforms vary across the market areas. The third one, the beachhead approach, enables a flexible usage of a platform, which can cover a performance range and several market areas.
The usage of these platform strategies enables reasonable cost structures in order to react on the increasing complexity of the planning and development activities due to shorter life cycles and broader product portfolios (Pil and Holweg, 2004). The overall objective is to get a maximum differentiation between the cars while sharing the maximum number of components (Weber, 2009, pp.4-6). The general main advantages of using platforms in form of standardized building blocks are represented in the shortening of the lead times in developing new derivatives through synergies, in the reduction of product development costs and in increasing scale effects concerning procurement and manufacturing (Mufatto and Roveda, 2000). Furthermore, the product reliability and quality increase through the usage of more mature components. Concerning management activities, platform strategies reduce managerial complexity. In best case, the external variety is increased by providing regional customized derivatives based on a shared platform, while the internal variety is decreased. Additionally, the usage of platforms fosters the business strategy flexibility by scaling the platform up or down and therewith enables an easy adoption to new market segments and requirements.
In order to operate a multi-platform strategy, bringing all the mentioned benefits, it is necessary to plan the number, allocation and components for the platforms (Jana and Grunow, 2016). The platform planning at automotive OEMs covers four central planning decisions. Firstly, the optimal number of vehicle platforms used (number of platforms) has to be clarified. Secondly, the optimal allocation of vehicles to platforms (vehicle-to-platform allocation) for an existing set of vehicles has to be determined. Thirdly, the optimal number of vehicles for each platform (platform portfolio), assuming there is no existing set of vehicles, has to be evaluated. Fourthly, the optimal components and modules to comprise a platform have to be selected for every introduced platform (platform selection). The vehicle-to-platform allocation and the platform portfolio decision are competing because, assuming a certain use case, either the set of vehicles is existent and given or not. Therefore, only one of these decisions can be implemented in an actual planning process.
The integration of platform strategies into the planning landscape and the four mentioned decisions have an impact on nearly all planning processes. Firstly, they affect the logistic and production processes. The operations planning resulting out of the platform decisions has an enormous impact on the complexity, operations, investments and overall costs (Mufatto and Roveda, 2000). Secondly, the development process of the platform itself has to be planned and operated. This step is already included in the reference process in Section 2.2. As the development of the platform contributes to the final development of the actual vehicle and its derivatives, the development lead time, standardization, quality and reliability are important influenced factors. Thirdly, the impact of using platform strategies also concerns the overall project organizational structure, i.e. teamwork, design task partitioning and new or additional supplier relationships have to be triggered and integrated into the existing planning processes. Furthermore, this covers the knowledge exchange among different projects. Fourthly and finally, the usage of platforms has a huge impact on the product characteristics and therewith on the customers’ perception (Jana and Grunow, 2016). The construction of a vehicle on a certain platform fosters commonality and therewith limits differentiation. These two factors can have a huge impact on the customers’ perception of the vehicle’s characteristics and pricing. In order to include these aspects into the planning landscape, especially implications on the demand planning have to be respected.
The strategic operations tasks, preponderantly introduced in the preceding Section 2.2, consist of the process technology design, location planning, strategic ramp-up planning, strategic procurement planning and have to respect implications of the demand planning.
As mentioned previously, process technology design defines the boundaries for the future manufacturing costs and feasibility by developing new process technologies. Recent examples are represented in the development of new manufacturing technologies for alternative powertrains or lightweight body design (Peters, 2015). In general, the design of manufacturing technologies is determined around fifty to six months before the SOP (Bornschlegl at al., 2015). The most important objective is to create sustainability concerning the manufacturing technology. This includes a reasonable cost structure, first main factor of sustainability. Most of the life cycle costs for the manufacturing technology are defined in an early planning phase. The main cost drivers are about development, investments, commissioning and training, operations, maintenance and additional elements like disposal. While some of these only occur once, others occur continuously. Concerning the operations costs, the focus on energy consumption gains an increasing importance, because of the customers’ growing ecological awareness and the overall pressure for CO2 emission reduction. Especially the energy costs can vary, depending on the country of location. The second main factor of sustainability is the production of high quality vehicles in a reasonable time. Altogether, production facilities represent the key factor for efficiency and effectiveness on cost, quality and time. They have to solve the conflicts between complex technical dependencies and various other objectives such as productivity, flexibility and availability, e.g. the available space can influence the production technology.
The task of location planning determines the production network structure in selecting suitable production sites and evaluating the construction of new plants at new locations. As described in Section 2.2, its main task is the platform-to-plant allocation and furthermore the allocation of production stages to plants (Jana and Grunow, 2016). The planning horizon for the location planning is between three and twelve years (Becker et al., 2016). Nowadays, as the complexity of production networks increases, the alignment of facility locations and expansions with the resulting flows in supply, production and distribution within the location planning stage is crucial (Kauder and Meyr, 2009). In order to react on unexpected or unforeseen events in the future, like a demand increase or a manufacturing shortfall, the construction of a robust and flexible network is necessary. Therefore, the overall objective is to maximize flexibility while to minimize cost. The conflict between flexibility and cost is considered the main tradeoff in location planning. This conflict even has an influence on the strategic facility design, which also has to be partly considered in this planning stage, but on a rather small detail level.
Based on the location planning, the strategic ramp-up planning determines the vehicle-to-plant allocation, selecting the plants for each vehicle of a platform (Jana and Grunow, 2016). The planning horizon spans several years (Becker et al., 2016). Besides the precise vehicle-to-plant allocation, the strategic ramp-up planning covers four more detailed decisions. Firstly, the ramp-up timing precisely determines the start of each ramp-up process, the ramp-up duration and the ramp-up sequence of the products at the plants. These information are summarized in the platform life cycle plan, stating the sequence and timing of the production cycles. Secondly, the strategic ramp-up planning decides about the shape of production capacity over time, i.e. to determine the maximum production capacity at each point in time. Thirdly, it deals with the production volume. In order to enable reliable and lean planning, the production volume and utilization must be identified. Fourthly, it deals with the transportation volumes. In case a product is produced at more than one plant, the precise distribution to the different markets has to be evaluated.
The next task, the strategic procurement planning, is not integrated into the reference process introduced in Section 2.2 in form of an individual planning task, but has a huge relevance especially for platform planning. As one of the main cost driver for the vehicle unit costs, it is necessary to make a rough evaluation of the potential supplier and the expected component prices. Therefore, it is helpful to establish long-term supplier relations, in order to get an accurate estimation of sourcing costs (Fleischmann et al., 2015). Furthermore, first strategic make-or-buy decision can be made and procurement requirements can be stated (Jana and Grunow, 2016). The main objective is to minimize procurement costs, while meeting the required quality standards.
Lastly, the demand planning needs to be discussed. Even though is does not represent an actual operations task, it has a high involvement in the other described aspects. Overall, the demand planning provides the basis for nearly all planning tasks, because the planning effort and the available budgets depend on the demand for the potential new product. The higher the demand for a certain product, the more investment costs are allowed to accrue, because the costs per unit will fall. As stated in Section 2.3, the usage of platforms has a further impact on the demand. It fosters commonality and limits differentiation. Hence, the usage of platforms can result in a loss of performance characteristics or even of quality characteristics (Bhandare and Allada, 2009). The resulting impact on the customers’ perception will affect the demand and pricing. Therefore, it is essential to include the demand planning into the planning of operations issues.
The platform strategy provides the basis for the planning of new vehicles and design projects (Jana and Grunow, 2016). Therewith, a fixed platform concept defines boundaries for the vehicle concepts, the network structure, the applicability of new process technologies and cost issues. Potential changes to the platform concept after the beginning of planning the design projects will result in a loss of synergy effects and additional costs for adaptions will occur. Therefore, a platform freeze, i.e. to definitely determine the platform strategy in order to initiate design projects, is crucial. This freeze has a long-term and irreversible impact of more than ten years. Hence, potential platform alternatives need to be discussed in advance of the freeze, including the analysis of implications on the detailed vehicle concept and customers’ perception. An insufficient evaluation of alternatives and wrong decisions can finally result in higher unit costs and investments, lost sales, usage of obsolete technologies or to customer rejections due to misguided demand anticipation. The evaluation of the alternatives requires the analysis of the impact of the platform freeze on the operations tasks of process technology design, location planning, strategic ramp-up planning, supplier network planning and includes the market perspective in form of demand planning and pricing. These aspects need to be aligned with and integrated in the cyclic task of platform planning.
As introduced as an important linked factor, the demand in form of the market perspective immediately effects the platform planning. The market perspective evaluates the customer perception, resulting in a demand forecast. The demand forecast itself includes pricing conceptions. Once fixed, a pricing concept influences the target costing, which defines the monetary boundaries for the platform planning and the further operations tasks.
Concerning the process technology design, the influence on and from the platform planning is limited, because this design project is usually already initiated in advance of the platform planning process. However, the selection of the platform is restricted to feasibility constraints and has a huge impact on the respective manufacturing life cycle costs. Therefore, the operations costs, investment costs, maintenance costs and development costs need to be integrated into the decision making process of platform planning. Quality factors and various other cost types can be complemented to this list.
Furthermore, network decisions about facilities and suppliers also have a huge impact on the cost and flexibility, as explained in the previous chapter. Hence, it is necessary to integrate the production network planning, including potential new sites, platform to plant allocation, allocation of production stages to plants and supplier network planning into the platform planning.
In addition, the integration of the strategic ramp-up planning is important in order to integrate production flexibility and cost into platform planning. The required alignment covers, amongst others, the product to plant allocation and the production capacity planning.
In summary, it can be said that it is necessary to integrate the mentioned operations tasks into strategic platform planning in order to enable a high degree of flexibility, reliability, and efficiency. Therewith a cost reduction and an early mitigation or avoidance of risks can be achieved.
Several models and literature sources concerning the decision making process of platform concepts already exist. Based on the concept design, the market perspective and cost effects, they determine potential platform concepts. In total nine different platform planning concepts are analyzed in the context of this thesis. An overview is given in Table 1. The different authors are listed on the left side of the rows, while the columns provide information about the included operations aspects and decisions. While the entry “o” declares that a certain aspect is integrated in an author’s paper, “-“ indicates that it is not integrated. The final point “flexibility” represents the opportunity to react on unforeseen events and act adaptive in general, e.g. to increase the production volume if necessary.
The concept design, as the first key aspect discussed in literature, is the main driver for the product design and its attributes. Thereby, the vehicle architecture is defined in functional attributes, determining the internal functions, and design variables, which define the parts and assemblies (De Weck, 2006). It provides a basis to cluster certain variants on diverse platforms and is discussed in various examples.
While AlGeddawy and ElMaraghy (2013) propose a heuristic to create modules and potential platforms based on cladistics with the objective to maximize product commonality, Bhandare and Allada (2009) generate a more complex approach. They introduce an intense definition and evaluation of design variables. Therewith, they provide a feasible solution in respecting technical constraints and finally determine common design variables, merging into a platform. Furthermore, De Weck et al. (2003) use a product variant design plan, which optimizes the product variants. These are stated in a design vector, which is limited to the platform design vector and is oriented to an objective vector. Seepersad et al. (2000) proceed similarly, but determine certain ranges for the design variables in dedicated constraints with an upper and lower bound. De Weck (2006) expands these approaches and formulates a tri-level problem. On the lowest level, the individual variants are formulated with respect to the engineering performance, which provides feasible boundaries for the objective values in form of upper and lower bounds. On the intermediate level the platforms are formulated and on the highest level the variants get assigned to the different platforms. A further approach by Kumar et al. (2009) applies a binary string, including the platform selection parameters, to define the product architecture.
Table 1: Overview of modeling approaches dealing with platform planning
Abbildung in dieser Leseprobe nicht enthalten
The second main task discussed in literature is the market perspective. The demand represents one very important indicator concerning target costing. Therefore, the customer perception on different attributes of a product and therewith on pricing is crucial in order to establish a reasonable price-cost relation. Concerning platforms, the usage of these can result in a loss of diversification and therewith in a reduced product price or lost sales.
Respecting these factors, Kumar et al. (2009) establish an approach based on a market segmentation grid, containing information about the existing markets and the competition in every niche. They use it in order to create a demand model. This is solved by a choice simulator, which makes use of a nested logit model. Furthermore, De Weck et al. (2003) propose a model, which calculates the potential demand volumes of the new products by comparing their performance against the one of the segment market leader. A different approach is established by Morgan et al. (2001), who integrate constraints about market segments and customer preferences in their modelling approach. De Weck (2006) proceeds similarly. Further papers by Bhandare and Allada (2009) or Seepersad (2000) assume certain demand scenarios as given.
Following on the market perspective and the target costing, the actual occurring costs and their relevance as a contributor to platform decision making is discussed in the existing literature as the third main aspect. The effects on costs are incorporated in nearly every of these literature examples, as cost reductions are one of the main benefits of platform strategies. The most common approach to integrate cost aspects is to immediately integrate them into the objective function in order to minimize costs or maximize profit.
Ben Arieh et al. (2009) introduce an example for a model, minimizing the costs for the platform production as well as the adding and removing of components to the platform. Kumar et al. (2009) even introduce the effects on warranty and repair costs. Another observed aspect is the learning effect introduced by De Weck (2006), who implements it in form of decreasing variable costs over time.
Further contributions to the operations perspective are only made sporadically. In Seepersad et al. (2000) the integration of the factors lead time and inventory costs is discussed in their model. Furthermore, De Weck (2006) integrates the required number of plants and plant capacity. In a similar way, Morgan et al. (2001) provide detailed capacity and lot sizing constraints. Besides, Bhandare and Allada (2009) address the product quality by implementing a quality loss, which is based on the Taguchi loss function, due to the usage of platforms.
The existing modeling and optimization approaches for the planning of automotive platforms mainly deal with the product architecture and the customer perception in order to optimize the platform strategy. Within the considered literature sources, a high degree of complexity is predominant. All of the sources determine the number of required platforms and, except from Kumar et al. (2009) and Morgan et al. (2001), all even determine the vehicle-to-platform allocation, whose implementation is preferred to the platform portfolio decision in the examined literature. Besides, in five out of nine times, the models support the platform selection and therefore increase their complexity significantly.
Concerning the cost perspective, the literature discusses various cost drivers, but not to a sufficient extent. Especially the development costs are barely discussed and incorporated. More general, the life cycle costs of a process technology and the respective process technology design are not integrated, even though their alignment is crucial to secure cost advantages. Furthermore, for a successful planning approach, a reliable quality policy has to be established. Quality effects are discussed in Bhandare and Allada (2009) and Kumar et al. (2009), however, a more detailed analysis of potential risk and further quality issues is necessary but not conducted in current research.
The impact of the actual operations planning task of location planning and strategic ramp-up planning is not evaluated at all to a mentionable extent. Even though Seepersad et al. (2000), De Weck (2006) and Morgan et al. (2001) address the aspects of lead time, capacity planning and lot sizing, respectively, the focus on operations planning has to be strengthened. This research need basically spans across all strategic operations tasks and decisions to be integrated into platform planning, covering, amongst others, the network planning, platform to plant allocations, capacity planning and supplier planning.
The integration of these operations aspects into the planning of new individual vehicles, beyond the platform planning, has already been researched to a significant extent. Fleischmann et al. (2006) propose a model, optimizing the allocation of various products of the BMW Group to global sites. This planning includes the supply of material, distribution of finished goods, required investments and the financial impact on cash flows. The vehicles get assigned to the different plants with respect to flexibility, capacity, product specific investment and structural investment, which include the planning of new production sites. Furthermore, the physical distance to the different markets is another decision component, which influences the transportation costs. The approach by Kauder and Meyr (2009) expands the model by Fleischmann et al. (2006) with further flexibility constraints, connecting all vehicle-plant chains. Therewith, it combines high flexibility with cost efficiency. A third approach by Becker et al. (2016) proposes to maximize the net present value of profit, including the discounted revenues, investment expenditures, production and transportation costs. In setting constraints about ramp-up and ramp–down timings, production capacities and the market perspective in form of demand and prices, they set boundaries for the optimal solution, while assuming products and production sites to be existent and known. These three models already cover most of the required operations aspects that are important within the process of platform planning.
However, current research neglects the implications of these aspects on platform planning so far. Therefore, it is necessary to align the tasks of process technology design, location planning, strategic ramp-up planning, strategic procurement planning and demand planning with the platform planning of automotive vehicles. In the following sections, a MILP based model is developed that covers the integration of various operations aspects into automotive platform planning. The objective is to determine the number of platforms and vehicle-to-platform allocation for an automotive OEM, while minimizing occurring cost.
The following MILP based model is established in order to integrate the operations perspective into the strategic planning of automotive vehicle platforms. It covers the platform decision of the number of platforms and the vehicle-to-platform allocation. Moreover, it incorporates different aspects of process technology design, location planning, strategic ramp-up planning, the market perspective and various cost drivers. In detail, it is based on a feasibility analysis, providing the possible combinations of platforms, vehicles and production sites. Furthermore, it takes into account life cycle costs concerning the process technology and the entire manufacturing process. These are influenced by a learning effect, depending on the quantity of vehicles that are produced on one platform. A dynamic timing approach enables the consideration of ramp-up timing and platform life cycles. Therewith, potential effects on the time-to-market aspect are respected. Besides, the model states a more complex view on the location planning, including the choice of existing or new production sites and the platform to plant allocation. This gets expanded to a flexible capacity planning. As the choice of platform impacts the customers’ perception of the vehicle, the following approach furthermore includes the market perspective, based on a demand forecast and indicating lost sales as a consequence of certain decisions. A coherent and uncommented version of the following model can be found in Appendix A.
Consider a firm that is about to develop the cost optimal assignment of various vehicle derivatives to its potential and available platform concepts. Therefore, the set of vehicle derivatives being considered is assumed to be existent and known. Hence, this modeling approach does not provide information about opportunities to place further derivatives on a certain platform. The same applies to the platform concepts. All platform concepts or scenarios are assumed to be existent and known in order to enable an immediate assignment of the vehicle derivatives to the platform without any further qualitative feasibility analysis. Furthermore, the model does not depend on a changing demand forecast, but assumes these data to be accurate and existent as well. Concerning the operations perspective, existent and potential new sites are known and thoroughly evaluated concerning all kind of occurring current and future costs. In addition, for the ease of use, it is presupposed that no vehicle derivative-platform combination is currently produced and that the vehicle derivatives need to be produced at the same site the prospective corresponding platform is produced at. Furthermore, at any site, there is only one production line for any production step, meaning that it is possible to produce different vehicles on the same production line without any further complications. Besides, newly installed capacity is immediately available and directly ready for use. All further planning data are determined deterministically.
Based on these assumptions, the cost optimal vehicle derivative-platform assignment can be determined by using a MILP based approach, which covers seven decisions and the corresponding decision parameters. Firstly, the integer variablexpismtdetermines the production volume of vehicle derivativei, which is produced on platformpat sitesfor the demand satisfaction of marketmin time periodt. The indexpis an element of the set of potential platform conceptsP, whileirepresents one vehicle derivative out of the set of all existent derivativesI. The site indexsstands for one certain production site out of all available or potential new sitesSof the production network. Besides, the indexmindicates the physical market locationmout of all marketsMthat need to be served. The indextdisplays the considered time period of the entire planning horizon T. Secondly, the integer variableksjtdecides about the capacity expansions at sitesfor the production stagejin periodt. It does not cover the construction and setup of new sites. The index of the production stagejis an element of the set of production stagesJ, which include the platform manufacturing, the press shop, the body shop, the paint shop and the final assembly. Thirdly, the binary variableypicorrelates withxpismtand determines the vehicle derivative-platform combinations. In case a certain derivativeiis produced on platformpit will turn one and otherwise stay zero. Fourthly, the binary decision variablezpiststrongly resembles toypi, but is expanded by the site location and the period. Therewith, it shows whether a certain derivativeiis produced on platformpat production sitesin periodt. In that case, it turns one and stays zero otherwise. Fifthly, the binary location decision variablefstindicates the usage of sites. In case a certain sitesis used in a certain periodt, it will turn one for that period, in case it is not used,fstwill stay zero for that site and period. Sixthly the binary variableesjtwill turn one in case the capacity at production sitesfor production stagejgets expanded in periodtand stay zero otherwise. Seventh and finally, the binary variablelpnestablishes learning effects for the manufacturing costs. In case a corresponding quantity bordernis exceed for the number of vehicle derivatives produced on one platformp, it will turn one and otherwise turn zero.
In order to describe the model, it is furthermore necessary to define numerous parameters in advance. These are required to formulate the precise objective function and constraints to set boundaries for the scope of solutions. At first, the seven different cost parameters are introduced. While states the overall procurement and manufacturing costs for the production of the platform p itself at production site s and representing the same costs in form of an average across all sites for every platform, the parameter declares the cost of necessary adjustments to vehicle derivative i, in case it is produced on platform p at site s. Moreover represent the costs of transportation for one vehicle derivative unit from production site s to market location m. Furthermore, the cost parameter specifies the costs to add a capacity unit to production stage j at production site s and states the required costs for a new production site s. In case the site already exists, these costs are zero. Finally, cSET represents the general setup costs for every derivative-platform-site combination. This includes start-up costs, commission and training. Besides the cost parameters, various others are required. While dimt declares the demand for vehicle derivative i at market location m in period t, the flexibility parameter sf sets a certain value smaller than one in order to set a maximum acceptable production utilization. Its counterpart, umin sets a certain value smaller than one and smaller than sf in order to set a minimum level of utilization for all production sites s that are in use. Besides ccpij includes the capacity consumption per vehicle derivative unit i for all production stages j, in case it is produced in platform p. In combination with that, CAPsj represents the available capacity at production site s for production stage j. It does only include the initially available capacity, but not the capacity expansions. The maximum additionally installable capacity at every production site for every production step is formulated in Ksj. In addition, the parameter BIG simply sets a sufficiently large number and Opist includes information about the feasibility, saying whether a certain platform-derivative-site combination is feasible in a certain period or not. According to the index n, CRBn includes the quantity boundaries that need to be exceeded in order to realize learning benefits and CRn represents the corresponding level of cost reduction. Finally, MINexp specifies the minimum required capacity expansion, in case the capacity gets expanded at any site s, LSpimt provides information about the lost sales volume in vehicle units for market m for period t in case a certain vehicle derivative i is produced on a certain platform p, and MARpi represents the general profit or marge for a sold unit of vehicle derivative i constructed on platform p. In fact, within MARpi a penalty fee is added to the actual marge in order to avoid the preferring of lost sales to the actual production of a vehicle derivative.
The overall objective of this MILP based approach is to assign the vehicle derivatives to a platform concept, while minimizing costs and respecting further operations factors.
Abbildung in dieser Leseprobe nicht enthalten
The objective function is stated in equation (1), containing eight different parts that influence costs. The first part determines the overall costs to produce the actual platforms themselves with their corresponding quantities. Based on a modular platform concept it is assumed that, if allowed due to later on explained feasibility constraints, a vehicle derivative can be manufactured on the basis of nearly every platform concept by just making some adaptions in form of removing, adding or exchanging certain components. These additional costs to adapt the platform to the derivative are formulated in the second part of the objective function. The third part describes a potential reduction of costs in form of learning benefits. In case a certain number of vehicle derivatives, which is high enough, is produced on one platform across the entire planning horizon, a learning benefit in form of a percentage of the average platform costs will be deducted from the overall costs. While the fourth part determines the transportation costs of all produced quantities from their production sites to the assigned market location, the fifth part covers the costs for additional capacity. Following on that, the next two parts incorporate the costs for the construction and installation of new production sites and the set-up costs for every platform-derivative-site combination into the objective function. The set-up costs will only occur in case the distinct platform-derivative-site combination is not manufactured in the previous time period, but it can occur several times for the same connection. On the other site, the construction costs for new sites can only occur once. The final cost part does not represent actual costs, but lost sales. Due to the choice of a certain platform, the customer perception of the product might change. This will result in a decreasing demand and smaller profits. Therefore, the lost sales due to platform usage have a negative impact and are considered as factual costs.
In order to set realistic boundaries for the scope of solutions, the following constraints take into account the operations limitations for the platform planning problem.
Abbildung in dieser Leseprobe nicht enthalten
The first constraint, equation (2), ensures that the demand will be fulfilled. Therefore, the produced quantityxpismtof derivativeifor marketmas to be as huge as the demand for the exact same vehicle derivative at the same market place for every time periodt. This initial equation gets expanded by the deduction of the lost salesLSpimtdue to the use of platform concepts, which may result in a decreased demand. In case the choice of platform has a negative impact on the customer perception, only the new and smaller demand has to be satisfied.
Bachelorarbeit, 56 Seiten
Bachelorarbeit, 70 Seiten
Facharbeit (Schule), 20 Seiten
Diplomarbeit, 148 Seiten
Seminararbeit, 25 Seiten
Seminararbeit, 28 Seiten
Hausarbeit (Hauptseminar), 32 Seiten
Hausarbeit (Hauptseminar), 24 Seiten
Hausarbeit (Hauptseminar), 23 Seiten
Der GRIN Verlag hat sich seit 1998 auf die Veröffentlichung akademischer eBooks und Bücher spezialisiert. Der GRIN Verlag steht damit als erstes Unternehmen für User Generated Quality Content. Die Verlagsseiten GRIN.com, Hausarbeiten.de und Diplomarbeiten24 bieten für Hochschullehrer, Absolventen und Studenten die ideale Plattform, wissenschaftliche Texte wie Hausarbeiten, Referate, Bachelorarbeiten, Masterarbeiten, Diplomarbeiten, Dissertationen und wissenschaftliche Aufsätze einem breiten Publikum zu präsentieren.
Kostenfreie Veröffentlichung: Hausarbeit, Bachelorarbeit, Diplomarbeit, Dissertation, Masterarbeit, Interpretation oder Referat jetzt veröffentlichen!