The Rise of Digital Manufacturing and the Boundaries of the Firm
This chapter discusses developments in the impact of new advanced materials within the auto supply chain, largely driven by new fuel and safety standards. It includes a brief overview of the new engineering tools that enable suppliers to undertake component engineering with the new materials. Then we lay out case studies of a steel company, a hydroformer and a moldmaker. One was historically a supplier of a commodity product, sheet steel. Another focused on building stamping presses. The third moved from traditional turning to combined machining and 3D printing. All are now involved in the design stage of products. The central theme is the combination of advanced materials and software. This is our definition of digital manufacturing.
Such suppliers challenge the traditional boundaries between auto OEMs and suppliers in the design and manufacturing stages of the product development cycle.
New Digital Tools for Manufacturing
Digital design tools, including CAD, CAM, CAE, FEA, and CFD, are already in use by many industries. They include:
- CAD (Computer-Aided Design) has almost entirely eliminated blueprints in vehicle design and engineering. Digital designs can then be fed into software to predict behavior and test innovative concepts. Designs can be quickly modified on the basis of such simulations, and almost instantly transmitted between suppliers and customers.
- FEA (Finite Element Analysis) performs linear, nonlinear, thermal, and dynamic analysis by treating an item as an assemblage of very small cubes. Modern computers can then calculate heat flow and mechanical stress and deformation using each cube’s basic material properties. FEA simulations can reveal thermal hot spots that will degrade performance or areas of stress that exceed safe material limits.
- FEA is now powerful enough to test laminates and composite materials, to check clearances with adjoining parts and even the ease of assembly. It can also simulate how a part will deform under stress, how much stress welds experience and so on, which allows “virtual” safety analysis.
- CAE (Computer-Aided Engineering) can optimize and validate product behavior before manufacturing, linking an engineer to previous engineering designs and material parameters. Specialized simulation software and validation test results can help them gauge whether their digital designs are likely to prove robust.
- CAM (Computer-Aided Manufacturing) can turn CAD files into programs that operate milling machines to make dies and fixtures. Programs can also simulate the assembly process, including how far a worker needs to step and their arm motions, to make sure there is sufficient physical room to undertake a task and to improve ergonomics.
- CFD (Computational Fluid Dynamics) can perform air and fluid flow analysis, including how an airbag expands or the movement of oil in a piston and the amount of heat it transfers. They can simulate the flow of gases and combustion process inside a cylinder. These calculations are more complex than those of FEA.
Many other software systems are employed in factories, to generate a production schedule, to automatically order parts and materials from suppliers and manage logistics, to reconfigure machines automatically, to monitor machine performance and individual part quality, and to allow remote reprogramming of robots and equipment.
As discussed in previous chapters, reduction in vehicle weight and drag to meet fuel efficiency standards requires the use of new lightweight materials. In addition to high-strength steel and aluminum alloys, which as materials are relatively homogenous in many automotive applications, these new materials include composites whose properties are less well understood. Simulation tools come to the fore, as their use requires careful control of the microstructure and an ability to understand how these affect material properties.
According to the Steel Market Development Institute Roadmap, with the dramatic increases in CAE capabilities over the past three decades, automakers have been increasingly relying on analytical performance validation of their design in preproduction phases. Manufacturers still validate physical parts and undertake early builds to check that performance matches digital models. Overall, however, the ability to reliably predict crashworthiness, durability and noise vibration and harshness ahead of hardware builds and physical testing has had a profound impact on product development cost and time.
Select Vignettes: Digital Manufacturing Across the Supply Chain
At the top of the food chain are the parts producing OEMs such as Delphi, Federal Mogul, and BorgWarner. They employ the leading software platforms extending across the full range of product life cycle management (PLCM) and also develop proprietary tools and modules to run within these platforms.
Delphi Software Package
The Energy and Chassis Systems is one of five Delphi divisions. CAD software was originally developed for the defense sector. Now much of the auto industry relies on CATIA, a product of the French aerospace firm Dassault. GM had acquired the similar software firm Unigraphics in 1991, but it was eventually spun off and since 2007 has been owned by Siemens. Given Delphi’s origins as the parts manufacturing operations of GM, they continue to use the Siemens Solid Edge® CAD software as the solid modeler of choice for machine, fixture, tooling, and gauge design. Delphi uses their NX® Software for plastic injection mold and sheet metal stamping dies. These are both part of the Siemens PLCM suite that help Delphi have a single tool to track projects from inception to the incorporation of specific parts on individual cars. As software improved, Delphi saw their machine control programming needs reduced by 70 percent. One advantage of remaining with the suite of Siemens products is the elimination of data translation, and greater ease of tracking design updates and reviewing designs from the concept stage through to manufacturing.
To increase productivity throughout the design to manufacturing cycle, Delphi developed Delphi’s Design Methodologies (DDM) for CAD/CAM design purposes, for which they received a PACE Award in 2004. By using a set of standardized methodologies for how to develop CAD drawings, designers can produce models that are easier for others to edit and change. As with other firms, Delphi patented pieces of their software approaches, and they are part of the intellectual property managed by Delphi Technologies Inc. (DTI) discussed in Chapter 7.
Automotive companies have completely outsourced many of their tool-and-die needs, which were highly specialized and subject to capacity constraints when the firm had many new products entering production. Now, software tools have allowed Delphi to move some of its mold making back in-house. They have transitioned to both high-speed machining and shop-floor programming in the company’s Flint, MI “29 Mold Build Toolroom”. A dedicated NC (Numerical Control) CAM application for the manufacture of complex shapes starts with Delphi’s CAD files. The software optimizes machining strategies and tool paths and so reduces the need to turn to outside machining specialists.
Federal Mogul
In 2007, Federal-Mogul Corporation adopted industry standard software for strategic PLCM. This makes it possible to carry out product development across the firm allowing operating units to control data and processes. Manufacturing engineers need to ensure that process plans, manufacturing bill-of materials (mBOMs), and work instructions reflect the current engineering design. It was a frequent source of problems in the days of blueprints, for example when different engineers made separate changes on their own copies of a drawing. PLCM helps assure that everyone works from the same version and updates are tracked and documented. This can then interact with manufacturing process management (MPM) so that the specification of the production process can be done in parallel with engineering, with an obvious savings in development time. Similarly, a documentation-tracking application uses prebuilt, workflow-driven project and document templates to help ensure that cross-functional processes are in accord with prescribed standards so that nothing falls through the cracks. It also formalizes critical quality management methodologies, such as Six Sigma, advanced product quality planning and adherence to the ISO 9000 industry standard for quality processes required by Federal Mogul’s customers.
BorgWarner
The BorgWarner Emission/Thermal Systems group is a leading manufacturer of emission control and cooling systems for automotive applications. For product design, BorgWarner uses an advanced 3D mechanical design system because of its capabilities for leveraging legacy data, its ability to handle polymer moldings and aluminum castings, and the availability of numerous add-on modules and specialty applications. In addition, BorgWarner added a module for collaborative data management. Another product area is BorgWarner’s TorqTransfer System for all-wheel-drive vehicles, which is electromagnetically actuated to provide varying levels of torque between the default front-wheel-drive mode and the rear axle. A potential customer needed a prototype in three months to meet their vehicle launch schedule. BorgWarner’s engineering software allowed them to reengineer their concept design for that firm, and demonstrate its reliability and proper functioning across a wide range of operating environments. At the same time, their engineers had to provide guidance for the design, tooling, and validation of the manufacturing processes. These and other tools, such as multiphysics software, help them bring much more robust products to market.
Most Auto Innovation Now Happens in Software
Virtual models allow leading Tier I suppliers and their automotive customers to reduce development time and costs, improving product quality. They link to CAM systems that use numerically controlled machines and robotics for the prototyping and validation phase, providing master model geometric data that can be transmitted to and from customers for evaluation, and tracked using PLCM tools.
Lower-tier firms typically have more limited software capabilities, but are also focused on a narrower range of activities. For example, an SME hot-forming company uses CAD in the manufacturing process to develop component prototypes. Another manufacturer that develops production systems for chains uses basic CAD and CAE software to work with customer specification sheets, to check the dimensioning of joining methods and to control the construction and building of prototype tools. At a small sheet metal shop, its engineers have mechanical and design engineering capabilities to both develop and make tools using various CAD products.
In summary, CAD in some form is omnipresent in the supply chain, as are design, manufacturing and product management software tools specific to a firm’s particular needs.
Schulze, MacDuffie, and Taube (2015) discuss knowledge generation and innovation diffusion in the global automotive industry. They focus on the central role of OEMs in system integration and their resulting dominance over product architecture and supply chain dynamics. Software tools enable shifts in business models in the auto supply chain where traditional parts producers or contract manufacturers now offer “manufacturing capabilities” across the range of supply chain services, that is Research–Design–Manufacturing–Sales–Service–Recycling. Smaller- and lower-tier suppliers tend to employ only individual tools, but there are many specialized SME suppliers that use modules within PLCM platforms for their particular design, simulation, and costing needs.
However the results are uneven. Interviews by Warrian and others hint at cultural factors limiting SME development in these new directions. Many SME founders and innovators found success as hands-on managers delivering manufacturing micro-efficiencies. They scaled their businesses and achieved career progression based on sweating details such as the cost of drills and other consumables. Software simulations and systems level gains are not in their DNA.
Amino: Sheet Hydroforming—Bringing a New Technology to Market
Amino is an Ontario-based supplier of panels and parts to auto OEMs, an SME with under 500 employees globally, headquartered in Japan. In 2015 they were a Finalist for the PACE Award with a manufacturing process innovation. Today the manufacturing technologies for vehicle unibodies are traditional sheet metal stamping, hot stamping, roll forming, and high-pressure hydroforming. This firm uses low-pressure sheet hydroforming and is experimenting with dieless forming.
The Japanese parent company is a stamping equipment manufacturer. Their operation in Southwestern Ontario was launched by a Canadian engineer who had worked for the parent company in Japan and made their first entry into parts production. The Ontario facility’s initial business consisted of the production of low volume sheet metal panels for models such as the GM Solstice, and for “dually” (double-rear-wheel) heavy duty pickup trucks.
It is the sole North American plant that uses low-pressure sheet hydroforming, a technology they developed over the past 30 years. Standard hydroforming uses a high-pressure (15,000 psi) fluid to shape tubular material inside a die, in effect expanding it like a balloon. That allows forming complex shapes, such as door pillars and engine supports, that would otherwise require welding two standard stamped pieces of metal along very long seams. Regular stamping stretches a sheet of metal between a male and female die, again at very high pressure. Sheet hydroforming adapts a standard stamping setup, but uses a male die that pushes the sheet into a pool of water that is kept at only a few atmospheres of pressure (under 100 psi). That results in a smoother material flow that allows very deep draws in a single cycle, even when it is an aluminum sheet. Furthermore, regular stamping requires a high-precision female die, and the pressures mean that they require expensive and hard-to-form steel. For parts that require more than one cycle because they are too deep for one pass because the metal would split, the multiple male–female die pairs can easily run several million dollars and take months to machine. Low-pressure sheet hydroforming incurs far lower capital costs, because it only uses one primary die and it can use softer, easier-to-machine materials. The drawback is that cycle times are lower, installed capacity is limited and many automotive engineers are unfamiliar with the technology.
The 2015 PACE candidate project was a sheet hydroformed clamshell liftgate for a Ford Lincoln SUV. Normal manufacturing for such a deeply curved piece would require two stampings that would be welded together. They were able to turn it out as a single piece. Close collaboration between Amino and Ford design engineers was essential. OEM engineers have computer models for standard stamping technology, but they are not adapted for the different metal flow characteristics of low-pressure sheet hydroforming. Amino had developed such models and was able to convince Ford that they were accurate and, with the normal sorts of modifications for angles and relief cuts, could turn out the liftgate that Lincoln stylists wanted. The ability to provide design capabilities using their own, proprietary software was key to the project’s success.
The company continues to engage in advanced development work on other metal-forming processes. One example is “Dieless NC Forming.” At a conceptual level this uses the millennia-old process of hammering sheet metal against an anvil to form complicated shapes without needing a die. (In practice, they use a rotating stylus on top and a small-diameter fixture below.) This process uses numerically controlled machinery and so can go directly from a CAD drawing to their forming machine.
A key challenge for innovative SME’s is attaining a viable minimum scale. Car companies are unwilling to design body panels to use low-pressure sheet hydroforming when the installed base is very small. But until firms see it being used widely, they are unwilling to themselves invest in creating a production line. The SME firm’s resources are too limited for them to build a second plant in the hopes that additional production business (or sales of their specialized presses) will follow. As a small firm, they also are limited in the number of customers with whom they can work, which leaves them at the mercy of the ups and downs of those firms—their initial customer had been General Motors for their Pontiac division, which was eliminated in GM’s 2009 bankruptcy. There is also their own particular history. They are a small, third-generation family business that manufactures sophisticated stamping presses, and were not naturally positioned to be a supplier of high-volume automotive parts.
This challenge for innovative SME suppliers is accentuated by the auto industry’s new geography that is based on global production systems, traced in Chapter 4. For example, the Ford Focus (sold under names that vary in different markets) is a standard vehicle produced in 100 locations around the globe. Its supplier partners are expected to co-locate in every region if they are to receive contracts for any region. In the case of the hydroforming company, they are a Japanese supplier with one production location in Canada, a small operation in China, but no presence in Europe, much less other developing markets. Small suppliers based in Europe or North America face the same underlying challenge. Such firms may be highly innovative, but face many hurdles in bringing their ideas to the global industry.
Plastics Injection Molding Case
This is a case study of ‘Tom,’ an entrepreneurial engineer whom we leave anonymous. He grew up in Oakville, the home of Ford Canada and trained as a machinist. He worked at McDonnell Douglas aircraft and then Husky Injection Molding, but their plant closed under competition from Chinese injection molders. He then ended up in Windsor, across the river from Detroit. It is the home of Canada’s leading edge tool, die and moldmakers, with 300-plus firms, more than Stuttgart, Germany. Eventually Tom started his own firm, using the highest quality, German machine tools—he is the lead user in Canada for DMG Mori. As he puts it, “I am a Mittelstand company. This is our approach to employees, technology and training”. His firm’s objective is to use state-of-the-art machining to make molds that enable a new level of precision in the final injection plastic products. This is important because, as with steel, plastics have changed from standard homogenous formulations of industrial polymers to the manipulation of microstructures.
With the slowdown in automotive, Tom entered the oil and gas pipe fixtures industry. The precision of his threaded products let him supply his customers in Alberta, from 2,000 km away. However, his core business came under pressure from Chinese injection molders as they gained technical skills and began to compete for mid-range automotive products. Where could his firm find a place in the shifting value chain?
In terms of technical demands, the plastics industry viewed molds for a coke bottle as the low end, a plastic auto part as the medium end and a plastic aircraft engine part the high end. The Chinese were taking over the middle range product. However, the rise of the new materials opened the possibility of entering the low end of the business in a new way, making the tooling for injection blow molds. (Figure 9.1)
Figure 9.1 Injection molding strategy: Leaving the middle to Chinese firms.
Bottling companies are under as much pressure as the auto companies in regard to environmental regulations. Lightweighting is imperative. However there is a key ratio of wall thickness to the height of the bottle. If the ratio is greater than 1:200 an error in the injection die of 0.001 inches can cause the collapse of the container. With the latest German machines his firm showed bottlers that he could maintain that level of precision for their molds. This led to a contract with the largest coke bottle maker in the United States located in Toledo, a short drive away. These are also the tolerances needed for the high-precision dies used in aerospace. The firm is now pitching their services to the Canadian companies in that sector, which are based in Montreal. Rather than trying to expand to become a global automotive supplier, the firm is using their ability to interact with customers to meet specialized needs to move vertically along the value chain.
Aside from the interesting biographical story, there is an important automotive supply chain point to be made. Academic researchers such as MacDuffie suggest that there will be innovative new players coming into the auto supply chain from the permeable boundary with related industries and technologies. Tom’s story suggests that there will also be movement of innovative suppliers out of the auto industry.
ArcelorMittal: S-in-Motion, Manufacturing for Design in Advanced Steels
Historically, the Canadian steel industry had a technically sophisticated, well-managed producer in Dofasco. They are now part of ArcelorMittal, but their engineering legacy persists. In addition, another firm that became part of ArcelorMittal, the French firm Usinor, had a strategic focus on making steels for the auto industry. Their technical center became the core of Arcelor Auto Design. As metallurgy became materials science, ArcelorMittal was well-placed to develop and produce new steels.
The challenge was to convince their automotive customers that these new, much more expensive steels were worth purchasing. If OEMs could be convinced to incorporate the new steels into their engineering tools, they might spot applications where weight reduction or additional strength were crucial, and be willing to pay a premium price. But their orientation was toward incremental changes to existing architectures. On the other hand, ArcelorMittal needed consistent and growing volume to justify the investment needed to make the new steels. They needed to demonstrate the benefits that could be had not from just substituting their new steels for an existing part, but in redesigning a unibody structure to take advantage of their ability to produce a wide array of steels of varying strength/hardness. To do so, would require developing the engineering capability to design and demonstrate the crashworthiness and weight savings of their new steels. The “S-in-Motion” project, begun in 2010, is how they have gone about that (Figures 9.2 and 9.3).
Through S-In-Motion, ArcelorMittal does full simulation analysis of the spring back of stamped parts, the forming of a unibody and performance in virtual crash tests. They combine this with prototyping capabilities that let them present potential customers with physical products. Because of their simulation capabilities, they can present customers with a design and accompanying engineering validation data that OEM customers can feed into their own engineering software platforms. Likewise, ongoing cooperation with manufacturers such as Cosma crosses another crucial divide, showing that the other parts of the supply chain can work with these new steels to meet quality, volume, and cost objectives. That allows customers to fit the use of such new materials within the development timeframe of a particular vehicle, from concept approval to design validation. In turn, PPAP (the production part approval process) must be finished before the SOP or start of production. ArcelorMittal’s engineering support can shorten the time to market from 3 years to 6 months or less.
Figure 9.2 The S-in-Motion Steel Body-in-White Project.
Figure 9.3 Extending the S-in-Motion Portfolio.
Engineering is the art of applying abstract principles to solve concrete problems. S-in-Motion needed to address that mindset. ArcelorMittal thus developed a demonstration vehicle—and now a series of 8 vehicles, from a full-sized pickup (2014) to a large sedan (2015), to a mid-sized SUV (2016)—that showcase the potential of the new steels, achieving weight savings of 20 percent or more over a “base” design. At the same time, S-in-Motion signaled that ArcelorMittal could provide validated engineering models that could let OEM engineers in the midst of a new vehicle program focus on the nitty-gritty sweating of the details. They would not need to delay a project while they convinced themselves and then their managers that a vehicle using these new materials would pass the various crash tests in a cost-effective manner.
The academic literature talks about a paradigm shift. In this case, the starting point is traditional Design for Manufacturing, where designers are expected to understand the practical limits of standard manufacturing methods and stick within those boundaries. The new paradigm is Manufacturing for Design (MfD), where the production side is tasked with surveying the range advanced materials and new manufacturing processes to contribute to the design being developed. S-in-Motion inserts itself into this process.
In the Honda door ring case in Chapter 8, Arcelor was able to take Honda’s vehicle concept and then devise the combination of new steels and production methods needed to turn that into a real-world product. That represents a transformation of steel’s place in the automotive design cycle. Into the 1980s the design cycle in automotive was 5 years for a new car. In the first year the team works with designers to see that a proposed style works with the new platform, which may include new powertrain elements and technologies coming out of advanced engineering. Tooling needs to be in place before the start of regular production, to assure that the processes are working as expected and to turn out preproduction vehicles for on-the-road testing and subsequent tweaks for noise, vibration, and handling. Creating tooling can take 9 months or more, so the design needs to be locked in place 12 to 18 months in advance, albeit there are many ways to finesse leads and lags. In the new automotive world where suppliers contribute design and engineering skills, they are brought in 2 and often 3 years in advance. By early 2017, most of the work for 2019 models will be largely complete, even if production facilities are not yet in place and kinks remain.
In the old days, steel was treated as a commodity. Automotive sheet steel was not necessarily a spot market, but steel mills were passive players, with the car companies making sure the mills had enough capacity while trying to lock in a good price. By the time they entered the picture, the design had long been frozen. Steel companies like ArcelorMittal are now seeking seats early in the development process. They need to get car companies to adopt innovative structural designs if they want to sell the new steels. To do so requires that they play the role of materials consultants to design teams from the inception of the design cycle.
The Honda door ring example in the previous chapter illustrates this. ArcelorMittal had developed the door ring concept as one element of their S-in-Motion design, and then “shopped” it to various OEMs. Honda took the concept and came back to ArcelorMittal with an early design for the door ring for the 2014 MDX. ArcelorMittal then took that design, and showed Honda how it could be realized using the tailored blanking of different steels. At the same time, they began working with their tailored blanking supplier, which happened to be one of their own divisions and not an outside firm, as well as Cosma, to develop a new welding process and to adjust the design based on Cosma’s experience with designing dies for hot stamping. All of this fed back to Honda engineers, who needed to be sure that the design would be ready in time—a variation of the roadmapping issue discussed in Chapter 7—and that it would work with the rest of the vehicle’s structure so as to earn a five-star Insurance Institute of Highway Safety (IIHS) rating. They were also concerned about weight and cost, but those objectives were met early on.
The use of the strategy embodied in the S-in-Motion unibody proved not to be a one-off success. OEMs found it fairly straightforward to use variations and it was quickly adopted for four vehicle programs in Europe and North America with cumulative projected output of nearly 3 million units. They now have tailored blanking facilities and steel production capabilities, by themselves or in partnership with other firms, in North America, Europe, Japan, and China. They have supplier-of-the-year awards from General Motors and Ford. Some 1,500 engineers showed up for a technology presentation at Hyundai in Korea. It is less visible, but this is all supported by growing their own automotive technical staff to 1,500 engineers across global design centers in France, Canada, and the United States. This enables them to respond quickly and to supply “resident engineers” who sit in the vehicle development centers of their customers and work alongside their engineers. While that is common for Tier I component suppliers, it has not been done in the past by steel firms (Figure 9.4).
Figure 9.4 Specific steel applications ArcelorMittal had developed.
Auto Innovation and the Boundaries of the Firm
The case studies in this chapter reflect the transformation of the supply chain and its implications for the locus of vehicle engineering. As the injection molding supplier case indicates, not all suppliers of technology find it profitable to exclusively remain in the industry, while SMEs such as Amino find it hard to break into what is now a global industry. The ability of steel companies such as ArcelorMittal, as well as their rivals in aluminum such as Novelis, to move from manufacturing into design has been more successful. Nevertheless, they are constrained by the product and platform architectures of the OEMs, which under a “platform” strategy use carryover parts and adaptations to previous designs rather than a “clean sheet” approach.
Innovation studies academics see these developments as an important lesson in the new global economy where the new digital technologies contribute to a mobility of production functions along global value chains. New business models arise as firms move forward, backwards, and sideways.