EOS: Growing the Business of Additive Manufacturing (AM)

Transcription

1 April 2017 A driver loses his leg in an accident and his doctor, instead of giving him a generic implant, scans his healthy leg in order to produce a prosthetic leg perfectly tailored to his body structure, greatly increasing his quality of life. A tool breaks in an assembly line and the workers print a replacement part in a matter of hours, eliminating the need to maintain an expensive stock of multiple spare parts or to shut down until a part is supplied. To some, these might have seemed like sci-fi scenarios in 2017, yet they were two real-life applications of EOS laser sintering technology, a form of additive manufacturing (AM). In September 2016, EOS was showcasing its new EOS M AM system, which used direct metal laser sintering (DMLS) technology, at Chicago s International Manufacturing Technology Show. It was EOS largest and fastest machine to date, capable of producing complex metal parts on an industrial scale and to industrial quality. However, despite the great strides that both EOS as a company and AM in general had made in the previous years, it still remained a relatively niche technology. In a machine-tool market worth $79 billion a year, 1 some $832 million worth of industrial 3D printers had been shipped in the first three quarters of Focusing on the high end of this market, EOS still faced significant challenges to convince industrial customers of the potential and reliability of this technology, which often involved rethinking the entire way a product was manufactured. 1 World Machine-Tool Output and Consumption Survey 2016, Gardner Research, April 27, 2016, last accessed November Global 3D Printer Market Up +35% in 2015 on the Back of B2B Purchases of Personal/Desktop Printers, CONTEXT, January 4, 2016, last accessed February This case was prepared by Professor Marc Sachon and Isaac Sastre Boquet, case writer, as the basis for class discussion rather than to illustrate either effective or ineffective handling of an administrative situation. April Copyright 2017 IESE. To order copies contact IESE Publishing via Alternatively, write to iesep@iesep.com or call No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means electronic, mechanical, photocopying, recording, or otherwise without the permission of IESE. Last edited: 4/25/17 1

2 Moreover, large technological corporations such as HP and General Electric had cast their eyes on the growth prospects of AM technology, having engaged in a flurry of acquisitions after announcing their entry into the market. Even though the marketing power of those companies would help to legitimize the technology, they would also become potent threats for EOS own growth. As the global leader in DMLS technology, EOS firmly believed in the future of AM and its many industrial applications. If the AM revolution was just around the corner, EOS wanted to be at the forefront. And for that, the company would have to find ways to grow even more. Additive Manufacturing (AM) Additive manufacturing was a manufacturing process where a component was produced by adding layer upon layer of base material (e.g., plastic or metal) until a 3D object was finally produced. The quality of the finished component produced by AM met or even exceeded the quality requirements for serial production. This additive approach contrasted with traditional subtractive manufacturing, where a piece of base material which had to be at least as big as the final part would be cut or milled, material gradually being removed from it, until an object was created. The vast majority of commercialized AM systems used the following workflow: 1) Computer-aided design (CAD) data are generated, describing the entire 3D geometry of the object. 2) This data are broken down into thin cross sections, generating a standard data file (usually in STL format), and fed to an AM machine. 3) The machine deposits the base material one layer at a time, recreating the geometry of the object. In the case of laser sintering as offered by EOS, a laser subsequently melts the material (metal or plastic) and the base material changes from powder to solid form. 4) After the object is removed, some postprocessing takes place (cleaning up, removal of support struts, etc.) until the object is finished. The first AM equipment and materials were developed in the early 1980s by Charles Hull, who built his first stereolithography ( 3D printing ) machine in 1983 and filed a patent for the process in However, even in 2017, AM was far from being a unitary technology. Rather, there were several processes with very different approaches that fit the general characteristics of AM. These processes differed mainly in the kind of materials that could be employed, the geometries for which they were best suited, and the type and amount of postprocessing that was necessary for the object to be completely finished after being removed from the build platform. These very diverse techniques could be grouped into three main families : 3 Extrusion: where a computer-controlled print head extruded a semiliquid material that, once solidified, formed the successive layers of the object. 3 Christopher Barnatt, 3D Printing: The Next Industrial Revolution, ExplainingTheFuture.com, IESE Business School-University of Navarra

3 P-1160-E Photopolymerization: where a material in liquid form was solidified by selectively using a laser or similar light source. Powder bed fusion: where materials in powdered form were turned into solid parts, using several possible processes, to form the successive layers of the object. All of EOS metal industrial 3D printers used a special form of this technology laser sintering. EOS focused on powder bed fusion, using melting as a way to fuse the powdered material. A thin layer of that material was applied to a base (the build platform), and the powder was selectively melted and welded together with a powerful laser beam, using the CAD data to direct the laser to the precise points that made up the geometry of the desired object. The platform was then lowered and another layer of powder applied, which was in turn fused selectively once more. The process was repeated until the full item was built. Then the part was removed, cleaned (unfused powder could be recovered and used to build another item), and finished. See Exhibit 1 for a depiction of the laser sintering process. Seat Belt Buckles 3D Print Source: Document provided by the company. AM processes had many advantages over other manufacturing techniques. First, AM greatly reduced waste, since most of the leftover material could be reused (unlike the metal chips produced in traditional processes). It also greatly simplified the manufacturing workflow, since geometrically complex objects including moving parts could be built in a single process. Moreover, it could build some geometrical structures that would be extremely difficult or even impossible to build using traditional methods, such as lattice structures that saved weight and material without compromising mechanical properties. Lastly, it was extremely flexible, since AM machines could produce Car suspension part 3D Print different parts without retooling or obtaining expensive new molds only a new CAD design had to be fed into the system. This greatly reduced the scale necessary to make short product runs profitable and enabled high degrees of product customization. Source: Document provided by the company. Its main disadvantages were the cost and variety of base materials available for manufacturing, the time it took to build a single item, and the size of the parts it could produce. It was still unable to build large, single-piece items (such as the tube of a drive shaft for a large vehicle). IESE Business School-University of Navarra 3

4 EOS EOS was founded in 1989 by Dr. Hans J. Langer, previously European manager at the laser positioning company General Scanning, 4 and Dr. Hans Steinbichler who went on to sell his shares to Langer in Its first customer was the BMW Group in Munich, which ordered the Stereos 400, EOS first stereolithography machine. EOS quickly went on to become the first major European provider of rapid prototyping systems. In parallel, the company started developing its laser sintering technology, introducing the EOSINT P 350 in In 1997, EOS sold its stereolithography division to 3D Systems, a company cofounded by Charles Hull, and focused exclusively on laser sintering after securing key patents. The 2000s saw the company expand its international footprint. It opened subsidiaries in the United Kingdom (2001), the United States (2001), India (2005), and China (2013), among other countries. In parallel, the development of EOS laser sintering equipment and base materials opened up new applications in fields such as the packaging, dental, medical, and aerospace industries. The company sold its 1,000th laser sintering system in 2011, having become the global market leader in that technology in EOS in 2016 Multimedia content EOS CEO Hans Langer discusses the origins of the company and some of the current applications it was working on Source: EOS EOS A Story of Success. Youtube.com. Last accessed March Illustration from istockphoto.com. At the end of 2016, EOS had an installed base of more than 2,400 systems in 52 countries worldwide. Its turnover of 315 million came from the sales of systems, base material (polymers and metals in powdered form), and consulting work and services. In 2014 it had opened a new 17,000 m 2 technology and customer center at the company s headquarters in Krailling, less than 20 kilometers from Munich, Germany, and had expanded its network of offices and distributors to 32 countries. (See Exhibit 2: EOS International Footprint.) By March 2017 yet another building had been added due to the rapid growth of the company. Figure A provides some basic metrics of EOS at the end of General Scanning merged with Cambridge Technology in IESE Business School-University of Navarra

5 P-1160-E Figure A EOS at a glance (2016) Turnover (millions of ) 315 Workforce 1,000 Of which: In Germany 70% In ROTW 30% R&D spending (% of turnover) 14% Source: Document provided by the company. Installed base (No. of systems) 2,400 Of which: Polymer systems 60% Metal systems 40% No. of customers with more than 1 system 428 Countries with EOS customers 52 Of which: North America 20% Europe & ROTW 60% Asia-Pacific 20% Moreover, EOS was in the middle of a period of very rapid growth. In the five years since 2010, the EOS workforce had more than doubled and, by March 2017, the company employed more than 1,000 people while revenues had grown by 400%. (See Exhibit 3 for EOS growth data.) The company s growth had taken off in 2010, coinciding with a reorganization of management and a repositioning of the company. Chief marketing officer Dr. Adrian Keppler, who joined EOS in January 2010, recalled the first meetings at the company: We crafted a vision. We said: We want additive manufacturing to be on the shop floor. We want to prove that this works. And working means not just one machine, not five, not 10. We want to see our customers producing hundreds or thousands of components, every day, with our machines. EOS had been able to fund this expansion through its cash flow, and the company remained in the sole ownership the CEO Langer and his family. EOS operated in accordance with ambitious three-year business plans. The plans of 2012 and 2015 had been successful. Consistent with this successful track record, the target for 2020 was to become a 1 billion company. However, EOS management was aware that it was at the beginning of a challenging journey. As Keppler reflected: If we want to position the company for serial production, we have to change the company. We have to think differently. We have to act differently. IESE Business School-University of Navarra 5

6 The Market for Additive Manufacturing Multimedia content Music played on a 3D-printed violin Source: Lisa Harouri Digital Forming: 3D Printed Stradivarius by EOS A Story of Success. Youtube.com. Last accessed March In February 2011, The Economist put AM on its cover with the headline Print me a Stradivarius, extolling the potential of AM as a manufacturing technology. It was one of the moments that EOS management recalled as showcasing how the technology was no longer seen as just a gadget for hobbyists nor as a tool limited to rapid prototyping. Indeed, as Keppler recalled: Before I joined EOS, I would talk with some colleagues and they would ask: What is industrial 3D printing? Now there are a lot of people aware of the technology and interested in what it can do for them. Several industries were thought to be great potential markets for AM. EOS had already been very successful in entering two of them: Aerospace: this industry was extremely sensitive to reductions in weight that could lead to fuel savings, or savings in the usage of very expensive alloys employed in key components. EOS had entered this industry very successfully. EOS machines could be found producing jet engine parts for the new Airbus A320neo as well as air ducts for Bell helicopters. Medical: the ability to build very complex geometries in a single process and the customization potential were the big draws for this industry. For example, lattice structures could be used to improve postoperative recovery and orthopedic replacements could be designed to match the patient s body. Furthermore, AMproduced implants were more lightweight, resulting in a better patient recovery and experience. AM was already extremely successful in the dental industry, with nearly eight million dental crowns, bridges and implants built every year using AM. (The EOS M 290 system could produce 450 crowns per day compared with the 15 to 20 crowns a dental technician could produce per day.) In 2016, EOS had 150 systems installed worldwide that were devoted to this activity. In addition to the medical and aerospace industries, there were several other industries with great promise for growth and in which EOS was already making inroads. In general, any industry that could benefit from AM s flexibility and customization in short production runs was a potential market for EOS. For example: Oil and gas prospection: this industry required very complex parts that had to be tailored to the characteristics of the drilling site and usually called for very short production runs. The ability to produce replacements for broken parts quickly was also key. Automotive: customization and flexibility in design were important drivers of future automobile sales. Companies such as Rolls-Royce and Formula 1 teams 6 IESE Business School-University of Navarra

7 P-1160-E already used AM for their highly specialized cars. Apart from customization, AM s ability to reduce the mass of components without compromising on performance made it the perfect manufacturing technology to lower CO 2 and other greenhouse gas emissions. The Challenges Railways/public transport: this industry was characterized by selling short runs of units (e.g., locomotives and wagons) tailored to the specifications of its B2B customers, plus spare parts. This made it an interesting market for AM given its ability to provide high levels of customization and low costs for short production runs. Lifestyle and sports products: from personalized bicycle parts to footwear perfectly tailored to the tastes and needs of the customer, AM could provide highly customizable items without being affected by short production runs. Other industrial applications: heat exchangers and resistors, for example, could be found in many applications from cooling down industrial machinery to cooling an item manufactured by injection molding. These were items with very complex geometries that were susceptible to more efficient production using AM. Replacement parts or molds could also be produced on demand by AM. EOS new flagship product in 2016, the M 400-4, was the largest and fastest system in the company s catalog. It could produce a metallic item with a maximum size of mm using a system of four lasers working in unison to provide a build rate of up to 100 cm 3 of processed material per hour. It was the pinnacle of the latest DMLS technology. It was not designed to be a rapid prototyping device but rather it was conceived and built for the factory floor, to produce hundreds of metal components every week. Improving the technology was just one of the many challenges that EOS faced in order to make the company grow in order to show that this works. Multimedia content Promotional video for the EOS M 400-4: 3D printing system for metal parts Source: EOS EOS M The ultra-fast quadlaser 3D printing system for metal parts. Youtube.com. Last accessed March Multimedia content Promotional video for the EOS M 400-4: Industrial 3D printing live Source: EOS EOS M Industrial 3D printing live!. Youtube.com. Last accessed March IESE Business School-University of Navarra 7

8 Moving Customers The first and most immediate challenge that EOS faced was to educate business customers about the potential of AM and how to adopt it in their operations. In general, EOS generated business through two channels: A proactive channel: where EOS targeted potential lighthouse customers and contacted them. These customers would light the way with their success stories, potentially establishing credibility for the technology and encouraging other companies to follow their lead. Selecting these customers and coming up with the best applications for them was one of the key strategic decisions of EOS. A reactive channel: where companies approached EOS asking for an AM application. This proposal could sometimes be precise ( Can you help me make personalized shoes? ) but sometimes it could also be very generic ( What can I do with AM in my company? ). Within the reactive channel, EOS had identified two sources of business generation: Bottom-up: a group of engineers, often on their own initiative, decided to push to introduce AM into their company s manufacturing processes, usually starting with a very limited budget and scope. The challenge was usually to convince management of the potential. Top-down: senior management, sometimes a visionary leader, would ask the company s engineering team to study and find ways to implement AM in the company s existing processes. The challenge here was to overcome the resistance of the established engineering teams, who would have to learn a new technology that might completely reshape their company s production processes. The sales process typically lasted 12 to 18 months from the moment the application was designed until it was fully implemented. Keppler explained, however, that the biggest challenge to getting a customer to adopt AM was a more strategic one: It is still difficult to generate business. It s one thing to be aware that there s this cool technology out there but understanding what I can do in my company with this technology, how to use it to add value to my core business that s a whole different matter. Companies generally had little know-how regarding AM, and the learning curve could be steep. As Güngör Kara, director of global application and consulting from the Additive Minds division of EOS, put it: What they would like is go to this or that place for two months and you ll know everything about AM. Well, that doesn t exist. To help customers bridge that knowledge gap, EOS had changed its approach, moving from being a system provider to becoming a comprehensive solution provider and adding a consulting arm to its business. It was tasked with assisting client companies to find the right application of AM in their companies and to provide the necessary know-how and engineer training. Still, finding this right application remained a big challenge, and Kara was aware that EOS would generally get only one shot at most at getting companies on board: Companies don t want to wait three, four or five years for an outcome that s not even given. 8 IESE Business School-University of Navarra

9 P-1160-E We have to make them successful otherwise, next time they will be blocked by their finance departments. EOS found that it was often successful in pilot phases or limited implementations but it was much more challenging to move from there to companywide implementations involving lots of engineers and a redesign of the company s production processes. As Kara put it: How do we go from four to 40, from 40 to 400? That s the next challenge. Moreover, making a good cost comparison study of AM and existing processes was not easy. Most companies tried to use the same ROI calculations that they used for their usual tooling and machine investments. However, it was not a like-for-like comparison when AM was involved. As Kara explained: For example, what happens if I take a component that previously was made of several smaller parts and that now I can build in a single process? Are there extra assembly costs, lead times, inventory costs that we are saving? How can we evaluate that? Gathering this kind of data is sometimes difficult, even for the companies themselves. Sometimes they themselves don t even know the real total cost they are paying to produce a certain part. Lastly, certification and validation were also an important hurdle in the way of adoption. Before serial production could be initiated, an AM system had to be certified and validated, ensuring that all parts produced had the required mechanical properties to be used reliably in the final application. This was a process that could take many months. Since every time that EOS introduced a new system it had to be certified, this made upgrading to more modern and capable systems a fastidious process. EOS engineers had to come up with ways of cutting down the time needed for validation. For example, the new EOS M used processes transferred from the highly successful and extensively validated EOS M 290. Competition EOS faced three kinds of competitors in the market for industrial-grade AM applications: Direct competitors that had come from the field of AM and had accumulated expertise similar to that of EOS in rapid prototyping before attempting to move to production. These included relatively large companies such as Concept Laser and SLM Solutions, as well as smaller start-ups. Large industrial companies that were making inroads into AM as an extension of their core engineering business. These included Trumpf and Renishaw. Large business-to-consumer (B2C) companies that had comparable business models and expertise that might be extended to AM. The most notable example was HP and its large 2D printing business. IESE Business School-University of Navarra 9

10 Figure B Global industrial/professional AM system sales (Q1 to Q3, 2015) Rank Company Units sold a Share 1 Stratasys (USA) 4,308 49% 2 3D Systems (USA) 1,938 22% 3 envisiontec (USA) % 4 EOS (Germany) 326 4% 5 mcor (Ireland) 265 3% Source: Global 3D Printer Market Up +35% in 2015 on the Back of B2B Purchases of Personal/Desktop Printers, CONTEXT, January 4, 2016, last accessed February a Includes printers priced at more than $5,000. The industrial/professional market was very heterogeneous, with systems ranging from $5,000 to $1.5 million in price. EOS was positioned in the high end of that market and only considered systems priced 150,000 and up as competitor products. This positioning meant that it sold fewer systems than many of its competitors but EOS systems were larger and more capable. EOS had a particularly strong position in metal systems, which were expected to be the largest drivers of growth in AM manufacturing in the coming years. This segment was also the priciest: the ASP (average system price) for metal systems was $465,000 per printer compared with an average of $115,000 for the whole industrial/professional market. 5 Figure C Global industrial/professional AM system sales, in dollars (Q1 to Q3, 2015) Rank Company Revenue a (millions of $) Share 1 Stratasys (USA) % 2 3D Systems (USA) % 3 EOS (Germany) % 4 SLM Solutions % (Germany) 5 Arcam (Sweden) % Source: CONTEXT, op. cit. a From printers sold for more than $5,000. The surge in AM growth had prompted the interest of large industrial groups, which were entering the market and flexing their considerable market muscle. Just as EOS was showcasing its newest system at the Chicago trade fair in 2016, General Electric had purchased two of the 5 Growth Slows in Metal 3D Printer Shipments in 1H 16, CONTEXT, November 15, 2016, Nov2016.pdf/f3a6399c-235a-4893-a fc29f512, last accessed February IESE Business School-University of Navarra

11 P-1160-E largest AM companies, SLM Solutions (Germany) and Arcam AB (Sweden), for $1.4 billion. 6 In May of the same year HP launched its Jet Fusion polymer systems for plastic 3D printing. This affected EOS in two ways. First, the influx of capital into the industry and the entrance of big organizations with lots of marketing power would help educate potential customers, develop the technologies and ultimately make the AM market grow. Second, companies such as GE and HP were tough competitors with vast resources and broad networks of industry contacts and they were masters of marketing. EOS believed that some of the company s strengths were its experience, technology base, and relationship with established manufacturing customers, particularly in Europe. EOS was confident that it was the only company in the market able to deliver a comprehensive solution encompassing all aspects of the process of building a component with AM, especially in metal. This core competence could be explained graphically with a triangle built around three vertices, all connected and influencing each other: - Systems: the AM machines ( industrial 3D printers ) themselves, which were very productive and created components of high quality. - Materials: in powder form that were processed to stringent quality standards and gave high-quality results in all EOS systems. - Process: the interaction between the laser beam and the material, which had to be completely mastered and controlled. Figure D EOS core competence triangle Source: Document provided by the company. 6 Rick Clough, Niclas Rolander and Andrea Rothman, GE Seeks to Drive 3D Printing Future With $1.4 Billion in Deals, Bloomberg, September 6, 2016, last accessed December IESE Business School-University of Navarra 11

12 EOS management considered that it had the best expertise in the industry across all three areas: system, material and process parameters. All three of these were intelligently harmonized to ensure the high quality of parts, so the EOS ecosystem could offer the most predictable, reliable, and reproducible results in terms of quality of the produced part. Given that AM was still a small and fragmented industry, no standards had yet emerged. In that regard, EOS had made its patents openly available for licensing. EOS was fully aware that it was competing not only against other AM companies but also against the same technologies it wished to replace in factories. It also had to provide answers for the new Industry 4.0 trends toward plant automation and cloud-based industry. Despite the strengths of the company, EOS management was aware that competition was going to get tougher. Keppler acknowledged that reality and the need to be ready: We have to have a different mindset. We have to be more aggressive, we have to sharpen our positioning, and we have to think harder about how to differentiate and leverage our network and ecosystem. Growing the Organization Since the new drive to get into serial production had been initiated in 2010, the headcount of EOS had grown threefold, from 332 to around 1,000 in January The company had expanded its structure, creating entire new units such as the Business Development Team and the Global Applications and Consulting Team. It had expanded its structure overseas, giving the subsidiaries more autonomy over the hiring of industry account managers with local industrial expertise. This presented a whole new set of challenges and tensions. Kara explained it graphically: This is a bit like teenagers. You grow very fast and suddenly your shoes no longer fit, or the shorts you always wore now look very strange. Therefore we need to align processes again and again and again. These growing pains were set to continue, as the new hires were key to the company s expansion. Keppler explained: The company comes from a background in rapid prototyping. We have a bit of a craftman s mindset and attitude. So we need to expand our know-how and support our existing people with new hires. We had to bring in a lot of new people, at different levels, from different backgrounds. And we have had to integrate them. EOS had a very strong company culture that came from the top. Langer, the founder and CEO, had established the company s core values in the 1990s, and those still remained. They were: Together: EOS wants to do things together, as a team. Responsible: everybody has to be responsible for his or her task, for his or her part within the team. Fair: everybody has to be treated fairly not only other members of the EOS workforce and management but also customers and suppliers. Excellent: day after day, all members of the EOS organization have to strive to improve not only themselves but the organization as a whole. 12 IESE Business School-University of Navarra

13 P-1160-E Finding the right cultural fit, with the right capabilities, in a tight recruiting market, was one of the biggest challenges EOS faced. As the company became more global, these problems would be compounded. As Keppler explained: What does our value responsible mean in the United States? Or in the Asia-Pacific? We are now a global organization and we have to ensure that all over the world we have a similar mindset. This globalization and the resulting need for autonomy in some parts of the organization represented a big change for a company used to a tight structure and strong top leadership. Finding a way to manage this evolution was central to the future. Moreover, one of the key challenges was that, in a growing company that was getting involved in more and more activities and businesses, what was previously a small and focused organization was now being pulled in many directions at the same time. Crossing the Chasm EOS wanted to lead the charge as AM aimed to bridge the gap between expectations and mainstream use. The company wanted to help bring about the kind of industrial transformation that it thought AM was capable of. To cross that gap and not fall in EOS knew that it had to build a critical mass and carefully ponder its moves, thinking of all the challenges ahead. Several questions arose: What was the best way to make the company grow? What was the best way to take the growth risks into account? How should the competition be tackled? Which were the best customers and markets to target? How should the competencies and capabilities needed to expand the AM market be acquired, without compromising the culture of EOS, which had brought the company to the position it was in to begin with? Thinking even further ahead, how could the company become faster, smarter, and more focused in order to tackle the challenges that lay ahead? IESE Business School-University of Navarra 13

14 Exhibit 1 General Functional Principle of Laser Sintering Source: Additive Manufacturing, Laser-Sintering and Industrial 3D printing Benefits and Functional Principle, EOS, last accessed March IESE Business School-University of Navarra

15 P-1160-E Exhibit 2 EOS International Footprint Source: Global Presence, EOS, last accessed March Exhibit 3 The Growth of EOS Source: Document provided by the company. IESE Business School-University of Navarra 15