EQUITY RESEARCH | April 13, 2016

Manufacturing is Daniela Costa undergoing its greatest +44(20)7774-8354 transformation since [email protected] Goldman Sachs International the Industrial Revolution. A wave of intelligent Yuichiro Isayama technologies is shaping a +81(3)6437-9806 [email protected] more connected, flexible Goldman Sachs Japan Co., Ltd. and efficient factory floor—and redefining the Joe Ritchie (212) 357-8914 ecosystem of equipment [email protected] providers in the process. Goldman, Sachs & Co. In the latest in our Profiles Tian Lu, CFA in Innovation series, we +852-2978-0748 examine the six [email protected] technologies driving this Goldman Sachs (Asia) L.L.C. transition and explore how William Turner it could yield more than +44(20)7051-0662 US$500 bn of cost savings. [email protected] Goldman Sachs International Yuki Kawanishi +81(3)6437-9804 [email protected] Goldman Sachs Japan Co., Ltd. Ashay Gupta (212) 902-9429 [email protected] Goldman, Sachs & Co.

FactoryPROFILESIN of theINNOVATION Future Beyond the Assembly Line

Goldman Sachs does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. For Reg AC certification and other important disclosures, see the Disclosure Appendix, or go to www.gs.com/research/hedge.html. Analysts employed by non-US affiliates are not registered/qualified as research analysts with FINRA in the U.S.

The Goldman Sachs Group, Inc. April 13, 2016 Profiles in Innovation

Table of Contents

Portfolio manager’s summary 4 What technologies lead the Factory of the Future (FoF)? 4 Opportunity for >US$500+ bn of cost savings 4 Where will these technologies be built? 5 Winners and losers in a changing competitive landscape 5 What could derail increasing adoption? 5 Factory of the Future in six charts 6 Why are we talking about the Factory of the Future? 8 Technologies: Bottom-up: >US$250 bn addressable market 14 Six innovative technologies for the Factory of the Future 15 Product Lifecycle Management software 18 Internet of Things Platform as a Service 20 Collaborative robots 23 Additive manufacturing 26 Automated Guided Vehicles 28 Radio Frequency Identification 30 Other key technologies (ordered alphabetically) 32 Industries: Top-down: >US$500 bn addressable market 34 Total addressable market could be worth more than US$500 bn 35 Winners & losers 45 New players, new business models 46 28 companies with exposure to the Factory of the Future 52 Drivers & barriers 65 Ten reasons why we are talking about the ‘Factory of the Future’ 65 What could derail it? 73 Disclosure Appendix 80

Prices in this report are as of the close of April 12, 2016.

This is the third report in the Profiles in Innovation series analysing emerging technologies that are creating profit pools and disrupting old ones. Access all the reports in the series below, and visit our portal to learn more and see related resources. Virtual & Augmented Reality: Understanding the Race for the Next Computing Platform Drones: Flying into the Mainstream

Other contributing authors: Robert D. Boroujerdi, Mohammed Moawalla, Jacqueline Du, Samuel Eisner, Willy Chen, Stefan Burgstaller, Wei Chen, Daiki Takayama, Heather Bellini, CFA, Simona Jankowski, CFA, Toshiya Hari, Diana Zhao, Gautam Pillai, Matthew Cabral, Jessica Kaur, Alex Karpos, Shateel Alam.

Goldman Sachs Global Investment Research 2 FACTORY OF THE FUTURE in numbers

UNREALISED POTENTIAL INNOVATIVE SOLUTIONS

The cost-savings opportunity The number of cost-saving automation from bringing manufacturing up- technologies we think will make the largest >US$500 to-date with the latest technology. impact on manufacturing. (p.15) 6 This equates to c.10% of current fixed investment. (p.35) billion The estimated size of the market for these technologies >US$250 by 2020, vs. the US$100+ bn The average potential cost-savings automation industry today. (p.15) per factory from introducing the billion latest automation technologies. (p.12) 10%-15% TECHNOLOGY IN PRACTICE COST-EFFECTIVE & COLLABORATIVE The increase in output per employee at Scott Fetzer Electrical Group after introducing The approximate amount of time it will cobots and automated guided take to recoup an investment in a 20% vehicles. (p.25) 1 vs. 3 collaborative robot (or “cobot”) in 2020 vs. the amount of time for an industrial years robot. (p.23) The number of buildings on Harley-Davidson’s York campus before and after adopting automated guided vehicles and PRIVATE INTEREST 41 vs. 2 robots to streamline production. (p.29) The increase in venture capital 188% investment in industrial software and robotics, respectively, over the past four The reduction in time-to- years. (p.72) market Maserati achieved using 139% product lifecycle management 50% software. (p.69)

MAN VS. MACHINE Current level of automation by industry (maximum of 100%)

Automotive 76%

Electronics 62%

Machinery 32%

Textiles, Leather & Apparel 25%

Wood & Paper Products 25%

Metals 21% Industries where we expect Food & Bev, Tobacco 18% “Factory of the Future” technologies to generate the Chemicals & Plastics 18% greatest cost savings (p.35) April 13, 2016 Profiles in Innovation

Portfolio manager’s summary

We focus on the key Manufacturing is entering a decade of significant transformation comparable to the emerging industrial revolution. While the years of abundant EM infrastructure investment are manufacturing gone, a growing middle class continues to demand higher wages, the skilled technologies that manufacturing talent pool is shrinking, and new global competitors have emerged. should allow significant Across most manufacturing industries, companies are faced with the challenges of cost savings, including much faster time to less demand and greater supply, putting strains on margins already challenged by market growing labour and transportation costs. This report leverages on our extensive interactions with companies and experts in the field globally (some of those profiled throughout the report). We identify the latest developments in manufacturing technologies that can materially improve cost structures, either through capacity utilisation or time to market. These should allow companies to maintain or improve their profitability and returns in an increasingly competitive environment. We also consider how these technologies might develop, reshaping the equipment makers’ competitive and regional landscapes.

What technologies lead the Factory of the Future (FoF)? We estimate the TAM of The concept of ‘Factory of the Future’ is a broad one. We narrow the scope of our the six technologies we analysis to three areas: (1) manufacturing design and production simulation – this analyse at >US$250 bn includes technologies such as product lifecycle management software (PLM) for both by 2020, around half factory and product design and Internet of Things software (cloud computing, data the US$500+ bn overall analytics); (2) physical manufacturing – industrial robots, collaborative robots allowing for potential cost savings human interaction (cobots), additive manufacturing (3D printing); and (3) in-factory we see as available to manufacturers (c.10% logistics – automated guided vehicles (AGVs), radio frequency identification (RFID). We of global fixed also take a brief look at adjacent technologies that might allow optimisation of manufacturing investment) processes or enable these six technologies (including virtual and augmented reality, machine vision, machine learning, nanotechnology and demand response). Opportunity for >US$500+ bn of cost savings Electronics, food & We estimate the total addressable market (TAM) opportunity for the ‘Factory of the Future’ beverage and capped by the potential cost savings in the key adoption verticals at US$500-650 bn machinery makers are (equating to c.10% of global fixed investment). We see the largest vertical the largest verticals for opportunities as electronics, food & beverage and machine making (e.g. tool makers, deployment electrical, resource and construction equipment). The TAM will likely be split between payments to firms making enabling equipment, and benefits passed on to customers to remain competitive. Margin increases for first-mover adopters are possible near term, but likely will erode as adoption spreads to the rest of the industry. We use two approaches to estimate TAM:

(1) Bottom-up (what is the revenue potential for equipment providers?): we assess the six key innovative technologies we think are most relevant in taking out stranded manufacturing costs, considering current penetration and potential pace of adoption to forecast the size of their markets. These six technologies add up to >US$250 bn by 2020E.

(2) Top-down (what is the cost savings potential for industries?): we break down fixed investment by industry and look at its level of automation (i.e. robot penetration) and digitisation (i.e. software penetration). We pro-rata the opportunity in each case, looking at the gap vs. the currently most-advanced manufacturing environment (automotive in robots; electronics in software), and at the average cost savings indicated for each industry by manufacturers of the key technologies. We estimate the TAM is capped at US$500-650 bn.

Cross-technology synergies, other technologies not profiled, deflating prices of technology as adoption broadens and, most importantly, benefits sharing between makers and users explain why our top-down forecast exceeds our bottom-up forecast.

Goldman Sachs Global Investment Research 4 April 13, 2016 Profiles in Innovation

Where will these technologies be built? Cost structures likely to We believe these technologies will allow manufacturers to be more agnostic over where converge regionally they locate their capacity, as cost structures become more similar across regions, with the over time main determinant increasingly location of demand. While EMs should be able to leapfrog stages of manufacturing infrastructure development (given less legacy physical capacity), adoption of the technologies detailed here will likely decrease the dependence of DMs on labour costs (still the main roadblock to reshoring of capacity in DMs).

Winners and losers in a changing competitive landscape See our profiles of 28 We expect to see tectonic shifts in the competitive landscape of equipment providers. The equipment makers number of players is rising, as software firms start to break into the industrial field, best-positioned to traditionally led by hardware players (e.g. Cisco recently acquired IoT platform provider benefit from Jasper). This will likely lead to new business models appearing (e.g. performance-based investment in the FoF contracts, equipment as a service). In addition, key breakthroughs in technology are (page 52) increasingly being driven by private, small firms. As a result, we expect more M&A within product categories, as well as regionally. We believe traditional capital goods incumbents with solid balance sheets (e.g. GE, Siemens, ABB) will engage in material consolidation of smaller players. Regionally, we see government support and the pressure from rising labour costs as driving Chinese companies to look to acquire international winners (albeit hurdles for deal completion remain high). We profile 23 listed covered companies further on in this report, and identify a non-exhaustive list of 24 private firms throughout this report.

What could derail increasing adoption? Standards, IP and Excess manufacturing capacity (fixed investment-to-GDP has been high for several years), legacy assets are key legacy asset bases, IP questions and a lack of standards for data transfer and compatibility barriers to adoption are the key roadblocks to wider adoption of these technologies, despite healthy corporate balance sheets to fund deployment. In addition, while technology cost has been falling, in several cases it is either still prohibitive, or the productivity gains achieved remain insufficient to materially change the user’s cost structure.

Exhibit 1: The cost breakdown of each manufacturing industry is different and therefore so is the technology impact Average cost structure by end market (EU-27) and machinery cost savings examples

Raw materials Labour Energy Services Construction D&A EBIT margin 100% Cobots allow >20% reduction in 11% 11% 12% 11% 10% 11% 9% labour costs which can equate to 90% 12% > 4pp of margin impact for an average machinery manufacturer 80% 22% 21% (e.g. Scott Fetzer Electrical 26% 24% 24% 22% 26% 22% Group case study) 70% 60% 15% 12% 15% 19% 22% 50% 19% 17% 21% RFIDs reduce labour costs >15% which can equate to >3pp margin 40% impact for an average machinery 30% manufacturer (e.g. Stanley Black 50% 53% & Decker case study) 20% 45% 42% 40% 44% 40% 41% 10% 0% PLM software that allows >10% reduction in raw material costs

Metals could equate to >4pp margin Textiles impact. Machinery Chemicals Electronics Automotive & beverages Food, tobacco Food, Woodpaper &

Source: Eurostat, Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 5 April 13, 2016 Profiles in Innovation

Factory of the Future in six charts

Exhibit 2: Labour costs continue to increase the pressure Exhibit 3: Customers are demanding unprecedented to automate levels of customisation and flexibility 5-year CAGR of annual manufacturing wages; local currency Combinations available when buying a Ford F150 pick-up

Theoretical 15.0% Equipment options Variants combinations 13.0% Trim 6 6 11.0% Passenger compartment 3 18 Power train 2 36 9.0% Cargo space 4 144 7.0% Engine 3 432 5.0% Transmission 3 1,296 Rear axle ratio 7 9,072 3.0% Wheels 9 81,648 1.0% Tires 8 653,184 Seats 18 11,757,312 -1.0% Power seats 2 23,514,624 US UK Italy Radio 5 117,573,120 China 5yr CAGR manufacturing5yr CAGR annual wages Japan India * India Russia France Taiwan Norway Canada Mexico* Sweden Belgium Estonia* Running boards 4 470,292,480 Australia Germany Singapore Rear windows 3 1,410,877,440 South Korea Colors 12 16,930,529,280 Developed market ● * Indicates where manufacturing wages Interior trim colors 3 50,791,587,840 were not available and average total ● Emerging market 16 individual options 12,870 653,687,735,500,800 wages were used instead.

Source: Siemens. Source: International Labor Organization, Haver, Trading Economics, Eurostat.

Exhibit 4: New, more affordable and flexible technologies Exhibit 5: Automation adoption rates vary dramatically are emerging as solutions across industries Labour costs and cobot installation costs, China Robot density per 1,000 employees 40 Automotive 80 - 120 35 Electrical/electronics 20 - 40 30 Plastic and chemical 15 - 35 25 products (inc. pharma) Metals 5 - 15 20

15 Machinery 5 - 15 Food products beverages; < 8 10 Tobacco products

5 Wood & paper products < 5 Textiles, leather, wearing 0 < 1 apparel Total manufacturing 15 -25 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 2031E 2032E 2033E 2034E 2035E

Annual temporary worker cost per person ('000s USD) 0 20406080100120 Co-bots annual cost per unit ('000s USD) Robots per 1,000 employees

Source: Goldman Sachs Global Investment Research. Source: Goldman Sachs Global Investment Research.

Exhibit 6: Six technologies are most relevant Exhibit 7: Top-down analysis suggests a >US$500 bn Summary of current level of automation, technology opportunity for those that become best in class penetration and relevance % fixed cost reduction vs. savings (mid-point)

Current level of Industries Current technology penetration Relevance automation 30% Manufacturing sub- Automation IoT PaaS PLM 3D Relevant Cobots AGVs RFIDs sectors index software software printing technologies 25% Food & beverage Automotive 76% ◕◕◔◑◔◑ All Textiles Chemicals 18% ◔◕○○○○IoT PaaS & PLM 20% Wood & paper Metals Machinery products Electronics 62% ◑◑◔◔◔◔ All 15% Machinery 32% ◔◑◔◑◔◑ All 10% Electronics

Food & Beverages 18% ◔◔○○◔◑All (ex. 3D printing) reduction cost % Fixed 5% Chemicals IoT PaaS, PLM & Metals 21% Automotive ◔○◔○◔○ AGVs

IoT PaaS, PLM & Wood & paper products 25% 0% ◕○○○◔○ AGVs 0 20406080100120140

Textiles & apparel 25% All (ex. 3D printing) Savings ($) ○◔○◔◔◔

Source: Goldman Sachs Global Investment Research. Source: Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 6

April 13, 2016 2016 13, April

Goldman Sachs Global Investment Research Exhibit 9: From design to servicing, the entire manufacturing ecosystem is becoming more connected and more intelligent Factory of the Future ecosystem

Profiles in Innovation Source: Goldman Sachs Global Investment Research. 7

April 13, 2016 Profiles in Innovation

Why are we talking about the Factory of the Future?

The current state of manufacturing Today, manufacturing accounts for 16% of global GDP. It is an important driver of an economy’s productivity growth, quite often the largest component of an economy’s foreign trade, and accounts for the largest share of private R&D spending in most major economies. Over the past two decades, companies have often relocated to EMs to take advantage of better growth and labour/environmental cost arbitrage. However, in the past three years, growth has faded, putting pressure on returns. This is driving companies to rectify key cost lines. Many factories are operating in a “disconnected” environment where the root causes of inefficiencies are rarely fully understood – automation and connectivity offer a solution to this. Current manufacturing mostly lacks cost efficiency in three dimensions that drive sub-optimal capacity use:  Machine-to-machine communication;  Two-way data transfer (from factory to enterprise resource planning system (ERP), but also from ERP to factory); and  Inter-factory integration.

While automation is already a US$100+ bn global industry, and robots are no longer science fiction, the penetration of semi-intelligent or intelligent production systems is still limited to a couple of industrial areas (autos and electronics). Automation is easier to achieve when a small number of products are made in large quantities. Major automakers and some other companies have already created production systems that are able to respond readily to these challenges, in terms of volume and product mix (e.g. Toyota has perfected Kanban manufacturing over many years and continues to refine it to this day; see Appendix 2). However, smaller companies and industries with end products that are not as large have yet to put such systems in place, and ample room exists for capacity and cost optimisation through broader technology adoption. The emergence of new technologies heralds an era of significant change in manufacturing. This is illustrated by the increased investment by venture capitalists (page 72) as well as large corporates. The opportunity allows some equipment providers to still benefit despite the trend of declining customer budgets (capex-to- sales), a characteristic of mature industries.

Ten reasons why we’re talking about the Factory of the Future Until the beginning of the last decade, factory automation was mostly triggered by the need for traditional manufacturers to cut costs while increasing productivity, in order to remain competitive with aggressive, lower-cost manufacturers entering their industries (e.g. US automakers such as GM in the 1980s fighting for survival in the face of intense Japanese competition).

Slowing growth recently, post 15 years of super-cycle, has seen a refocusing on manufacturing margins and returns. Since the financial crisis, returns have stagnated, as operating leverage has faded and fixed costs have continued to rise (mainly labour costs). We highlight ten reasons why we believe manufacturing will see a significant focus on implementing newer technologies over the coming years:

Goldman Sachs Global Investment Research 8 April 13, 2016 Profiles in Innovation

1. DMs need to reduce costs to remain competitive vs. EMs. The average DM manufacturing hourly labour cost is still more than four times higher than the average manufacturing wage in China.

2. But, rising EM wages are lowering the ‘low-cost location’ edge. For example, in China, wages are growing at mid-teens percentages vs. less than 3% growth in most developed markets.

3. Specialised manufacturing labour is increasingly scarce. The scarcity of specialised labour is exacerbated in the largest manufacturing nations by ageing populations (e.g. retiring baby boomers) and in EMs by younger generations that no longer aspire to work in the factories.

4. Productivity is increasingly a way to differentiate when legacy capacity is abundant. The last 15 years have seen an unprecedented level of capex globally. This drove a new set of competitors to emerge, and in turn led to unproductive capacity in many global industries, making time and cost to produce more significant differentiators between manufacturers than availability of capacity.

5. There is an increased push to shorten time to market, particularly by eliminating stranded inventories. Today’s market environment means information comes faster and is more accessible than ever, with customers now expecting products sooner. This is a key source of being able to generate a better return on capital for a manufacturer.

6. Customers demand unprecedented customisation. Personalisation today is used as a competitive tool to capture sales and is something customers look for to distinguish their purchases. In addition, the growth of consumers from emerging markets, which are made up of a diverse range of cultural and ethnic groups, increases the complexity of manufacturing in consumer segments when attempting to appeal to these new, large markets.

7. Focus on safety and security has increased dramatically. A more automated, controlled and less labour-intensive environment should reduce the likelihood of accidents and costly litigation. In addition, factories need to be safe from cyber-attacks, owing to the increasingly important role data has in the manufacturing process.

8. Several governments are actively pushing to stay ahead in manufacturing. In a globalised world, with lower trade barriers, scarcer demand and greater supply, countries have to fight hard to maintain competitiveness and to position themselves as locations of choice for manufacturing. We have seen numerous initiatives over recent years that stress this (e.g. China Manufacturing 2025, Germany’s Industrie 4.0).

9. Short-term demands from shareholders for dividends and buybacks put further strain on cash available for organic investments. Short-term macro uncertainty and a lack of visibility over growth has led to increasing demand from investors for cash returns from corporates (dividends and buybacks). While this might not be a sustainable form of long-term capital allocation, it has weighed on the short-term considerations of corporates and puts further pressure on optimising FCF generation through more productive capex.

10. Finally, key technologies now exist for fully optimised and connected manufacturing. There have been rapid improvements in the capabilities of a number of technologies which can drive substantial change in factories across the world (e.g. sensors, computing power and robots).

See the section on Drivers and Barriers (page 65) for more detail.

Goldman Sachs Global Investment Research 9 The Ecosystem Factory of the Future - Key Players A sample of companies with exposure to enabling technologies

MANUFACTURING DESIGN AND PHYSICAL MANUFACTURING IN-FACTORY LOGISTICS PRODUCTION SIMULATION Product Lifecycle Additive Manufacturing Automated Guided Management (3D Printing) Vehicles (PLM) Software 3D Systems Amazon (Kiva Systems) Ansys EOS Bosch Arena Solutions Han’s Laser Clearpath Robotics Autodesk Hewlett-Packard Daifuku Aveva HG Tech Dematic Dassault Systèmes Renishaw JBT Corporation Hexagon AB Stratasys Jungheinrich PTC Trumpf KION Group KUKA Internet of Things Collaborative Robots (Cobots) Nidec Platform as a Service ABB Group Savant Automation Advantech Automata Technologies Seegrid Amazon Fanuc Siasun Atos Foxconn Toyota Bosch Harmonic Drive Systems Cisco [Jasper] Kawada Industries Radio Frequency Cumulocity Kawasaki Identification (RFID) Device Insight KUKA Alien Technology GE LifeRobotics Cognex IBM Rethink Robotics CipherLab Ingenu Siasun DHL Microsoft [Azure] Teradyne [Universal Robots] FedEx MJ Intelligent System Yaskawa HG Tech Oracle Honeywell PTC Impinj Rockwell Automation Keyence Salesforce Smartrac SAP Zebra Technologies Schneider Electric Siemens The Ecosystem Factory of the Future - Sizing the Market

Total Addressable Market for Enabling Technologies (2020E)

US$230-285 bn MANUFACTURING DESIGN AND PRODUCTION SIMULATION US$30-35 bn Product Lifecycle Management Software Development stage: Commercialised Used to simulate the lifecycle of a product from inception to servicing to end of useful life

US$200-250 bn Internet of Things Platform as a Service Development stage: Early commercialisation Used for machine-to-machine communication and remote control of factories

US$28-34 bn PHYSICAL MANUFACTURING US$25-30 bn Additive Manufacturing (3D Printing) Development stage: Early commercialisation Used to create 3D objects from a digital model

US$3-4 bn Collaborative Robots (Cobots) Development stage: Early commercialisation Used to perform unstructured, often repetitive tasks alongside humans

US$4 bn IN-FACTORY LOGISTICS US$2 bn Automated Guided Vehicles Development stage: Early development Used to move materials throughout production and storage

US$2 bn Radio Frequency Identification Development stage: Mature with new applications Used to identify and track objects via radio signals

Cost-Savings Potential by Industry Technologies with greatest potential (as a % of Max cost-savings fixed PLM 3D Manufacturing vertical potential (USD) investment) IoT Software Cobot Printing AGV RFID Electronics $120 bn 13% Machinery $110 bn 23% Food & Bev, Tobacco $100 bn 27% Automotive $70 bn 8% Chemicals & Plastics $70 bn 7% Metals $70 bn 21% Wood & Paper Products $50 bn 25% Textiles, Leather & Apparel c.$10 bn 27%

April 13, 2016April 13,

Goldman Sachs Global Investment Research Exhibit 10: Our top-down analysis suggests a US$500-650 bn opportunity to bring manufacturing to best-in-class Industry-by-industry potential addressable market

Maximum Industries Current size of industry1 Current degree of automation 2 Potential savings 6 opportunity size % of global manuf. Robot intensity Robot Software % of Software Manufacturing sub- Automation Savings from FoF fixed asset Amount ($bn) per 1,000 penetration fixed asset penetration ($bn) sectors index technologies investment employees 3 index investment4 index 5 Automotive 19% 895 97 90% 9% 63% 76% 6 - 8% 50 - 70

Chemicals and 21% 990 19 19% 3% 17% 18% 5 - 7% 50 - 70 plastics

Electronics 20% 965 33 34% 17% 90% 62% 9 - 13% 90 - 120

Machinery 10% 485 8 9% 9% 54% 32% 17 - 23% 90 - 110

Food products, 8% 365 5 6% 5% 30% 18% 20 - 27% 70 -100 beverages & tobacco

Metals 8% 360 10 11% 5% 30% 21% 16 - 21% 60 - 70

Wood and paper 4% 200 2 2% 8% 47% 25% 18 - 25% 40 - 50 products

Textiles, leather & 1% 50 0 0% 8% 49% 25% 20 - 27% c.10 wearing apparel

Others / Not classified 9% 440 13 14% 7% 42% 28% 8 - 10% 40 - 50

Total Manufacturing 100% 4750 21 21% 8% 47% 34% 10 - 15% 500 - 650

Industries Pressure to automate Current technology penetration 7 Relevance factor

Manufacturing sub- Net margins Net margins 30yr average Current age (100 = very IoT PaaS PLM IoT PaaS PLM ∆ net margins % difference Cobots 3D printing AGVs RFIDs Cobots 3D printing AGVs RFIDs sectors (2010-14) 6 (2015-18E) age 7 (2015) likely) software software software software Automotive 4.9% 5.4% 0.5% 6.1 6.5 6.3% 85 ◕◕◔◑◔◑ 1.00 1.00 1.00 1.00 1.00 1.00

Chemicals and 7.0% 7.5% 0.5% 8.0 7.9 -1.4% 72 0.25 1.00 0.00 0.00 0.00 0.00 plastics ◔◕○○○○ Electronics 10.2% 10.0% -0.3% 7.5 9.5 27.2% 80 ◑◑◔◔◔◔ 1.00 1.00 1.00 1.00 1.00 1.00 Machinery 7.9% 6.3% -1.6% 9.3 10.5 13.1% 86 ◔◑◔◑◔◑ 1.00 1.00 1.00 1.00 1.00 1.00

Food products, 9.3% 9.9% 0.6% 8.5 8.6 1.4% 68 1.00 1.00 0.75 0.50 1.00 1.00 beverages & tobacco ◔◔○○◔◑ Metals 2.0% 2.7% 0.7% 11.8 11.4 -3.7% 72 ◔◔○○◔○ 0.75 1.00 0.00 0.50 0.75 0.50

Wood and paper 4.8% 7.0% 2.2% 8.0 9.2 13.9% 74 1.00 0.75 0.50 0.25 1.00 0.75 products ◕○○○◔○

Textiles, leather & n.a. n.a. n.a. 9.2 12.2 33.0% n.a. 1.00 0.75 1.00 0.50 1.00 1.00 wearing apparel ○◔○◔◔◔

Notes: 1 GFCF data from the World Bank, broken up using data from the OECD and China Statistical Yearbook 5 Automotive software penetration adjusted up according to analyst discretion 2 Robot penetration indexed against automotive; software penetration indexed against electronics 6 Net margins of global GS coverage in each industry, note not entire industry and some companies in the sample have exposures to other industries 3 7

Weighted average of density in Germany, Japan, France, Italy and UK Average age data is for the US only Profiles in Innovation 4 Software as a % of investments in private non-residential fixed assets in the US; automotive includes other transport equipment 8 Qualitative assessment of penetration at analyst's descretion (actual penetration may be less or greater)

12 Source: Goldman Sachs Global Investment Research.

April 13, 2016April 13, Goldman Sachs Global Investment Research Exhibit 11: Manufacturing has always been evolving

Cobots IoT software Autonomous in-factory logistics PLM software Additive Manufacturing (industrial scale) Nanotechnology ... Factory Wide Control Systems Artificial (extension to process industries) intelligence Functional Automation and Manual Mechanization of Mass Computarization (in discrete Work Production Areas industries)

Timeline Before Industrial Industrial Revolution 70s-90s 90s-2000s Next 5 years 5 - 10 years from now >10 years Revolution until the 70s from now from now

Winners Workers Large scale manufacturers Traditional Capital Goods Traditional automation Software providers 3D printer manufacturers Suppliers of complex chemical manufacturers manufacturers (industrial & traditional) and composite materials Industrial software providers Hardware makers (e.g., IT Industrial software providers Sensors & enabling technologies Software / AI providers suppliers) (e.g., ABB, Siemens, Invensys, Niche manufacturers Rockwell Automation) Robotics providers (e.g., ABB, Semiconductor manufacturers Kuka, Fanuc)

Losers - Unskilled workers Unskilled workers Unskilled workers Low-value hardware Low-cost country manufacturers ... Ultimately companies with providers (weak aftermarket) lack of ability to adapt due to Small size manufacturing Small size manufacturing Small size manufacturing Manufacturers of cutting either weak competitive shops unable to adapt shops unable to adapt shops unable to adapt Unskilled workers tools and machines, hand tools positioning or low cash (e.g., Sandvik) returns

Manufacturers of spare parts and supplies

Discrete automation manufacturers?

Capital goods companies mainly Can capital goods companies benefited from changes still benefit from here? Which ones will be able to use these technologies to lower their cost bases and time-to-market?

Source: Goldman Sachs Global Investment Research. Profiles in Innovation

13

April 13, 2016 Profiles in Innovation

Technologies: Bottom-up: >US$250 bn addressable market

 Technological advancement is enabling greater flexibility, productivity and connectivity within and among factories

 We see six technologies as most disruptive, falling into three groups:

Manufacturing design and production simulation

1. Product lifecycle management software (PLM)

2. Internet of Things: Platform as a Service (IoT PaaS) Physical manufacturing 3. Collaborative robots (cobots) 4. Additive manufacturing (3D printing)

In-factory logistics

5. Automated guided vehicles (AGVs)

6. Radio frequency identification (RFID)

Goldman Sachs Global Investment Research 14 April 13, 2016 Profiles in Innovation

Six innovative technologies for the Factory of the Future

The factory is evolving and is entering an era of new technological innovation. We see six technologies as key to this evolution: PLM software, Internet of Things (IoT) platform as a service, collaborative robots, additive manufacturing, automated guided vehicles (AGVs) and RFID. In the factory of the future, manufacturing should become increasingly flexible with seamless integration of a range of physical and digital systems, and devices communicating with each other to optimise production. We believe these technologies will allow the shift from mass-production to mass-customisation that is required in the next era of manufacturing.

We expect the entire production process to become more intelligent, from design to services:

1. Intelligent design: From inception, factories and products will be designed more intelligently using the latest modelling and simulation software, optimised to reduce downtime and with the construction process less likely to be subject to delays and complications. This is particularly crucial since we estimate that choices made during the design phase can affect up to 70% of the costs of a new product.

2. Intelligent administration: A remotely managed factory, allowing supervision of factories located in low-cost areas, while working from the corporate HQ. Supply chain management, allowing manufacturers to monitor and manage inventory in real-time, across different factories/geographies, “Industrie 4.0 will impact the whole reducing transportation and inventory costs. product lifecycle end to end – from design to production, the actual usage 3. Intelligent production: Smart factories phase until end-of-life – and cannot be improving operational performance attributed to one single department of through greater productivity and reduced the firm. The digital transformation is a costs. This is continuously being optimised cross-functional effort that needs to be addressed by the whole company.” using automation equipment and the data

it collects. Cloud factories might emerge, Dr. Reinhold Achatz, Head of Corporate enabling manufacturers, particularly new Function Technology, Innovation & entrants, to focus on design and sales, Sustainability at ThyssenKrupp rather than heavy factory investment. Corporation Smaller-scale production closer to customers is also a possibility.

4. Intelligent product: Big data of users’ experiences helps make product design adjustment/upgrades easier, as well as increasing customisation of products. Data emitted also helps generate after-sales.

5. Intelligent sales: Increased customer-to-manufacturer direct sales. Optimised sales/ advertising channels based on buyers’ prior purchases.

6. Intelligent after-sales: Predictive maintenance, resulting in more stable cash flow generation and avoiding unnecessary downtime for the customer, as well as unnecessary, costly after-sales personnel (see Kaeser Kompressoren vignette page 49).

The technologies we focus on in this report address each of these steps above. When taking a deep-dive into the practical equipment/software that allows these benefits to be realised, we cluster these technologies into the three areas discussed previously: manufacturing design and production simulation (1, 2, 4 and 6 above); physical manufacturing (2 and 3 above); and in-factory logistics, a part (although not all) of that encompassed in 2, 3 and 5.

Goldman Sachs Global Investment Research 15 April 13, 2016 Profiles in Innovation

A bottom-up approach to forecasting TAM for six core technologies We make a detailed assessment of the six technologies that we believe will be most disruptive to current manufacturing, grouped into three key areas: (1) manufacturing design and production simulation; (2) physical manufacturing; and (3) in-factory logistics. We forecast the TAM for each technology independently. For some, we rely on third-party research. For others, we estimate and take into account the following factors:

 The current status of the installed base where these technologies will be applicable (e.g. cobots replacing labour that is in short supply);

 The level of existing penetration of the disruptive technology (e.g. PLM software penetration is currently around 20%-30% according to Siemens, and currently only 1%-2% of in-factory vehicles are automated);

 We estimate the evolution of the penetration rates of these technologies over the next five years (e.g. we estimate AGV penetration to rise to over 15% in manufacturing). While many technologies in unconstrained conditions should follow the path of Moore’s Law, in reality, technology adoption will likely be driven by the following elements:

o The average age of equipment in each industry, as an indicator of the level of cumbersome legacy equipment (potentially incompatible with new technologies) that remains as a barrier to adoption;

o The regional pressures to automate/digitise, such as labour cost growth, constraints on skilled labour, government incentives, etc.; and

o The development of key technology enablers’ cost curves and value-add potential, in order to command a greater proportion of customer spend.

Clearly, technologies are adopted at unpredictable rates, and entirely unexpected applications can arise to dominate their uses. Furthermore, while we look at each technology in isolation here, many of these emerging technologies will be used in conjunction with each other, potentially multiplying their impact.

Exhibit 12: Age of legacy equipment should be a key Exhibit 13: Market readiness and maturity of key determinant of adoption of new technologies technologies Average age of manufacturing assets in the US; 1947-2014 Key technologies market readiness vs. technology maturity

10 100% Niche Deployment 9.5 90% High Sensors 9 80% 8.5 Industrial 70% robots Data analytics 8 Additive 60% manufacturing 7.5 RFIDs Digital 7 50% Cloud computing simulation 6.5 Pilot Barriers Age of equipmentin the US 40% Collaborative robots 6 readiness Market IoT PaaS 30% 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 20% AGVs Manufacturing Manufacturing median 10% Low 0% 0%Low 20% 40% 60% 80%High 100% Technology maturity

Source: Company data, Goldman Sachs Global Investment Research. Source: Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 16 April 13, 2016 Profiles in Innovation

FoF technologies as enablers of the Internet of Things The Internet of Things (IoT) concept is starting to revolutionise the manufacturing landscape. Equipment is becoming more digital and connected, forming networks of machines and new ecosystems. While we are still in the nascent stages of adoption, we believe that the timing of the transformation will resemble that seen in the consumer sector (e.g. penetration of mobile phones rose from 5% to 100% in 20 years), but that the economic implications will be much bigger (we have seen estimates of 2.5-5.0x the size of the consumer internet). See our report The next industrial revolution: Moving from B-R-I-C- K-S to B-I-T-S, July 16, 2014 for more. The technologies highlighted in this report both help enable the IoT and are critical to harnessing its full potential.

Less is more at Daimler

We interviewed Markus Schaefer, Member of the Divisional Board of Mercedes-Benz Cars, Production & Supply Chain Management, discussing what the company is doing to stay ahead of the competition in automotive manufacturing.

The challenge: Daimler realised at an early stage that rapid growth of car production, a greater number of products (ten years ago, it had 15 models; now it has 40), increased customisation and complexity of those products, and a rapidly changing business environment require a very flexible and efficient production network.

The solution: To increase its flexibility and ability to handle complexity, Daimler is investing in its global production network to standardise the plants and adopt the latest production technologies. The company is moving away from fixed conveyance production lines into smaller individual production cells. This more fragmented system will reduce the number of robots by increasing the utilisation of each, and also prevent the possibility of an entire production line having to be halted in order to reprogramme an industrial robot. The use of IoT networks and AGVs will allow for just- in-time transport between cells, quicker than a sequential production line. The process is also made more flexible through the use of technologies such as collaborative robots that can work alongside human operators; Daimler believes such technology will play an important role in the future of manufacturing. At the moment, it is working alongside robot supplier Kuka in particular, and currently employs 20 cobots in production. The company notes, however, that this technology first needs to overcome a number challenges, for example, to become more affordable and lighter, and that it requires regulatory and other approvals.

Some things don’t change: When asked how Daimler chooses a factory location, Mr. Schaefer reiterated that the most important factors are closeness to end markets, logistics cost, low labour costs and tariff incentives (beyond location of demand). Its arsenal of technologies will be deployed when appropriate, regardless of location.

Exhibit 14: The use of individual production cells and Exhibit 15: A Kuka cobot working inside a Daimler car AGVs makes assembly more flexible while humans work from the outside

Source: Daimler. Source: Daimler.

Goldman Sachs Global Investment Research 17 April 13, 2016 Profiles in Innovation

MANUFACTURING DESIGN & PRODUCTION SIMULATION

• Advanced computational methods to create a 3D simulation of entire Product Lifecycle product life cycles Management software • Commercialised • Savings example: MWV reduced time-to-market from 18 to 6 months US$30-35 bn (2020E base case) using Dassault Systemes’ PLM software

The increasingly complex nature of products and the manufacturing process of these, suggest that the future manufacturing process itself will be designed and perfected in a virtual design and simulation. Manufacturers can use advanced computational methods to create a 3D simulation of the entire manufacturing process, avoiding costly errors, optimising the efficiency of the factory, and dramatically increasing time-to-market. Sensors and detection hardware help to “synchronise” simulation and reality at every point in the production process. Using similar technology, manufacturers can create a simulation of the entire life-cycle of a product, from inception, through design and development, to servicing and disposal. Better designs are a pre-condition for applying new manufacturing technologies, realising full productivity potential and dealing with the indirect costs associated with increased technological complexity. Payback times typically range between one and five years, depending on the current status of the process implemented, the addressed problems and the maturity of the business.

Competitive landscape We highlight two key players: Siemens and Dassault Systemes, but there are also other major players such as Ansys, PTC, Autodesk, Aveva and Hexagon. These key players tend to have specialties, with only Dassault Systemes and Siemens having close to a complete end-to-end offering

The challenges Engineering software needs to be developed differently for each industry, and requires a learning ramp-up to use the software and to adjust engineering techniques. Another challenge is the constraint imposed by computational capabilities; the greater the computational power, the more simulations are possible and the more effective the software.

The markets This technology disrupts conventional 2-dimensional design and architecture. It disrupted dramatically reduces production errors and increases time-to-market. Therefore, engineers have to adapt and incorporate it or risk being left behind.

Applications and PLM software has historically been high-usage in complex manufacturing industries and is industries exposed now penetrating all industries. For example, Dassault Systemes focuses on four core industries: aerospace & defence, transportation & mobility, marine & offshore, and industrial equipment. However, it also now highlights new industries: consumer retail & packaged goods, energy, process & utilities, high tech, financial services, natural resources, construction and even life sciences.

Sizing the revenue PLM software has a potential TAM of US$30-35 bn by 2020E. We arrive at this range opportunity using estimates from Siemens and Dassault Systemes, the two market leaders in PLM software. Siemens estimates that the market has a medium-term growth rate of 8% pa and a market size of US$23 bn in 2015.

Goldman Sachs Global Investment Research 18 April 13, 2016 Profiles in Innovation

Siemens’ PLM software achieves a three-fold increase in process planning capacity

The challenge: Perkins, a manufacturer of diesel engines and power solutions, wanted to eliminate late-stage changes to expensive tooling and production processes. It also wanted to reduce the risk of escalating costs and disruption to production schedules.

The solution: The company introduced Siemens’ PLM software: Teamcenter and Tecnomatix for successful new engine launches; this achieved a three-fold increase in process planning capacity compared with its previous approach. It was able to significantly reduce its time-to-market and continually increase its unique customer sales configurations (the 1200 series had 100 configurations in comparison to less than 40 previously). Other benefits included tighter collaboration among colleagues, vendors and customers throughout the product lifecycle.

Exhibit 16: Siemens PLM software designs products… Exhibit 17: ...plants, processes and more

Source: Siemens. Source: Siemens.

Goldman Sachs Global Investment Research 19 April 13, 2016 Profiles in Innovation

MANUFACTURING DESIGN & PRODUCT SIMULATION • Smarter manufacturing, increased productivity, reduced Internet of Things downtime/costs • Early commercialisation Platform as a Service • Savings example: GoPro cut its freight costs by 75% & inventory costs by 9% by introducing SAP’s platform to manage the supply chain US$200-250 bn (2020E base case)

Internet of Things platform as a service (PaaS) encompasses software applications, cloud-based storage, big data analytics and an industrial internet operating system that serves as the platform. In our view, this will be a differentiated growth avenue for select industrial/technology companies that gain first-mover advantage. Already, a significant number of companies are spending more than 20% of their tech budgets on big data, and we believe this is about to accelerate. In our view, the next step is an integrated platform with horizontal applications (e.g. asset monitoring, predictive maintenance, dynamic manufacturing), like GE’s Predix, that allow for exponential data capture/analysis and real- time decision making. Competitive While Cisco is at the forefront of broader IoT trends, we see GE dominating within landscape industrials; others, including Rockwell Automation, Schneider Electric and Siemens, are also participating prominently. Importantly, traditional software/hardware suppliers, including Amazon (Amazon Web Services), Atos, Axeda, SAP, IBM, Microsoft, Oracle and Salesforce, will be key partners/competitors in this space.

The challenges Cyber security, legacy infrastructure and protocol standardisation: (1) as IoT software becomes more sophisticated, and more operations fall under the supervision of a holistic system, privacy, data security and network reliability become increasingly important concerns; (2) legacy infrastructure has been a bottleneck, but the exponential increase in connected devices has been an enabler; (3) there is a lack of defined standards for machine connectivity, which a platform like GE’s Predix aims to resolve.

Applications and Asset monitoring and safety/reliability checks have been around for over two decades, but industries exposed these were largely limited to either regular servicing visits by engineers (which help preempt downtime) or circuit breaker type checks built in to shut down a system in the event of a safety hazard. With IoT PaaS, companies can now: (1) predict when a malfunction is imminent; (2) perform real-time assessments to help improve performance; and (3) remotely adjust operations. In our view, this application is relevant to all industries.

Sizing the revenue IoT PaaS has the potential to be worth US$200-250 bn by 2020, we believe. Both GE opportunity and Cisco believe the market for cloud computing and IoT-based software/analytics will exceed US$220 bn by 2020.

Exhibit 18: Software penetration in fixed investment has Exhibit 19: …and in other countries been increasing for a while in the US… Software as a percentage of gross fixed capital formation Traditional capital goods equipment vs. software as a percentage of total investment in fixed assets; US

25% 12.0%

11.5% 20% 11.0%

15% 10.5%

10.0% 10% Formation 9.5%

5% 9.0%

% of total investment in fixed assets fixed in investment total % of 8.5% 0% Softwareof Gross asFixed% Capital 8.0%

1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Traditional Capital Goods Software France UK USA Japan

Source: US Bureau of Economic Analysis, Goldman Sachs Global Investment Source: OECD, Goldman Sachs Global Investment Research. Research.

Goldman Sachs Global Investment Research 20 April 13, 2016 Profiles in Innovation

Exhibit 20: Improving efficiency is one of the key benefits Exhibit 21: …and big data has an impact across the entire of the industrial internet… manufacturing value chain Survey of 250 industry executives Use of big data in the manufacturing value chain matrix

R&D and Supply-chain Marketing Aftersales Production How important are the following benefits in driving business to adopt the design management and sales services Industrial Internet? Build interoperable, cross functional R&D and product 100% design databases to enable concurrent engineering 

90% Aggregate and share customer data to improve service, increase sales, and enable design-to-value   80% Source and share data through virtual collaboration 70% sites (idea marketplaces to enable crowdsourcing)   60% Implement advanced demand forecasting and supply 50% planning across suppliers and use external variables 

40% Implement lean manufacturing; model and optimize production; develop dashboards  30% Implement sensor data-driven analytics to improve 20% throughput and enable mass customisation  10% Collect real-time after-sales data from sensors and 0% customer feedback to trigger services and detect flaws.  Optimize Reduce Improve Create new Improve Enhance Enhance asset operational worker revenue sustainability customer worker safety Improve supply-chain visibility through control towers and organisation-wide collaboration utilization cost productivity streams experience 

Extremely important Very important Important Somewhat important Not important

Source: World Economic Forum (2015). Source: McKinsey Global Institute.

Predix – GE’s leading Internet of Things platform as a service

What is Predix? GE’s Predix, a cloud-based, open-sourced platform, is designed specifically for the industrial user. Think of it as the industrials version of Windows/Android with horizontal applications (e.g. asset performance, brilliant manufacturing, etc.) that can be customised by industry. Predix is meant to be the world’s first and only cloud-based software platform built by and for the industry. Domain expertise and industrial internet capability are key areas that GE believes will differentiate its offering from more mainstream tech companies like Amazon (AWS), Microsoft (Azure) and Salesforce.

As an example:

The challenge: GE Aviation analysed 340TB of data from 3.4 mn flights on 25 airlines to help improve asset performance and minimise disruptions.

The solution: The result of implementing Predix: performance was boosted 287x and cost lowered 7x.

What does it mean for GE? We believe there is application beyond GE’s installed base, but see partnerships with traditional ERP players as critical in driving the convergence of information technology and operating technology. It is unclear who will ultimately own the customer relationship. In the end, with US$225 bn of value at stake by 2020E, we expect the industrial IoT to provide a robust, margin-accretive revenue stream over the next several years for early adopters such as GE.

Enabler: Cloud The cloud decentralises storage, managing and processing of data. This, in turn, results in computing a more productive and flexible way for companies to manage their IT, with the bulk of computational work and storage done remotely, allowing them to streamline their businesses and focus on their core competencies.

Enabler: Big data Previously, factory managers and workers gathered on location to discuss and resolve analytics issues in the production process. In the future, big data analytical software should allow for automatic identification and adjustment of the manufacturing process to monitor and improve efficiency.

Goldman Sachs Global Investment Research 21 April 13, 2016 Profiles in Innovation

Exhibit 22: Cloud-based software helps manage production within a single factory and across a network of factories

Analysis Analysis Production adjustment

Over production

OK

Internet server/Cloud OK Factory D Factory E Production OK adjustment

Over OK OK Under Factory B Factory C production production

Production Factory A adjustment

Source: Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 22 April 13, 2016 Profiles in Innovation

PHYSICAL MANUFACTURING

• Lower-cost, smaller and more flexible robots capable of working Collaborative robots alongside humans • Early commercialisation stage US$3-4bn (2020E base case) • Savings example: PLC Industries boosted its output per worker 40% using cobots supplied by Universal Robots

We see collaborative robots or “cobots” (robots that work alongside humans), as one of the fastest-growing areas in manufacturing machinery. The technology is typically smaller, lighter, cheaper and more flexible than traditional robots. Until recently, the robotics industry was dominated by the “big four”: Fanuc, Yaskawa, Kuka and ABB; but cobots open up the market to smaller players and start-ups, as well as increasing the economic viability of robotics in general industry. We estimate the payback time of a cobot to be less than one year in comparison to nearly three years for industrial robots. As the technology develops, we expect it to displace low-skilled, repetitive labour tasks and in the future, expect these robots to incorporate increasing degrees of self-learning and rational independent decision making. Key enablers, machine vision and machine learning, are discussed on page 32.

Competitive landscape We highlight Teradyne (Universal Robots), ABB, Kuka and Fanuc. Kawada and Yaskawa also have a collaborative robot offering. Indirect exposure is possible through suppliers such as Harmonic Drive Systems. There are also a number of start-ups and private companies in this space including Automata Technologies, LifeRobotics and Rethink Robotics (Baxter).

The challenges Standardisation, improving performance, safety, weight and cost. The potential for cobot use is huge, but first it must overcome a number of challenges: (1) there needs to be a standardised programmable platform to increase adoption rates; (2) performance (such as speed) must continue to improve; (3) they need to be safe and meet regulations as they will often be used in hazardous environments; and (4) the cost and capability of the robot must be sufficient for investments to be economically viable (currently US$40-90k).

Applications and Cobots still work more slowly than conventional industrial robots, and can only perform industries exposed simple operations. Therefore, we do not foresee them being used on highly automated lines requiring little human input. However, we believe cobot penetration could advance relatively quickly in areas where human involvement in production is needed, including distribution/conveyor lines and assortment lines that use AGVs, as well as flexible production cells. As the performance of cobots improves, we forecast a gradual increase in their use on food/medicine packaging lines and semiconductor assembly lines.

Sizing the revenue We estimate total sales of cobots of US$3+ bn by 2020 and >US$6 bn by 2025. Using opportunity data from the US, Japan, Western Europe, South Korea and China, we forecast cobot demand to have greater potential in countries with the following features: (1) a likely large decline in the manufacturing workforce; (2) labour costs exceeding/catching up to robot installation costs; and (3) a certain level of conventional industrial robot penetration and productivity. Subject to these and supply assumptions, we forecast the cobot TAM to 2025.

Exhibit 23: Cobot cost recovery period should be dramatically shortened over 10 years Cost recovery model for cobots (‘000s US$)

Co-bots 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E Average robot price 30.0 29.1 28.2 27.4 26.6 25.8 25.0 24.2 23.5 22.8 22.1 System Integration cost 60.0 58.2 56.5 54.8 53.1 51.5 50.0 48.5 47.0 45.6 44.2 Total cost 90.0 87.3 84.7 82.1 79.7 77.3 75.0 72.7 70.5 68.4 66.4 YoY, % -3% -3% -3% -3% -3% -3% -3% -3% -3% -3% Working hour per unit (hours) 24 24 24 24 24 24 24 24 24 24 24 Average labor cost 25.5 26.4 27.8 29.5 31.2 33.2 34.7 36.2 37.8 39.5 41.2 YoY, % 3%6%6%6%6%4%4%4%4%4% Working hour per man (hours) 8 8 8 8 8 8 8 8 8 8 8 Net staff replaced 22222222222 Depreciation saved 3.0 2.9 2.8 2.7 2.7 2.6 2.5 2.4 2.4 2.3 2.2 Maintenance costs 9.0 8.7 8.5 8.2 8.0 7.7 7.5 7.3 7.1 6.8 6.6 Payback period (years) 2.0 1.9 1.7 1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.9

Source: Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 23 April 13, 2016 Profiles in Innovation

Exhibit 24: Labour costs likely to exceed robot Exhibit 25: Cobot demand to emerge early in advanced installation costs by around 2020 in China economies China: Comparison of labour costs and installation costs Major advanced economies: Workforce and latent cobot demand forecasts

40 21,000

35 20,000 30

25 19,000

20 18,000

15 17,000 10

5 16,000

0 15,000 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 2031E 2032E 2033E 2034E 2035E 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E Annual temporary worker cost per person ('000s USD) Work force demand (thous) Co-bots annual cost per unit ('000s USD) Co-bots force (conversion into manpower) Temporary worker population (thous)

Source: Goldman Sachs Global Investment Research. Source: United Nations, Goldman Sachs Global Investment Research.

Exhibit 26: Estimating the potential of the collaborative robotics market US$ mn

Collaborative robots 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E Total worker population (thous) 1,471,313 1,470,389 1,468,401 1,465,884 1,463,566 1,461,897 1,460,635 1,459,805 1,459,161 1,458,189 1,456,496 Population ratio for manufacturing 23.5% 23.5% 23.5% 23.4% 23.4% 23.3% 23.3% 23.3% 23.3% 23.3% 23.2% Population ratio (temporary worker) 38.3% 38.2% 38.2% 38.1% 38.1% 38.1% 38.0% 38.0% 38.0% 37.9% 37.9% Manufacture regular worker population (thous) 213,622 213,340 212,883 212,342 211,838 211,446 211,170 210,970 210,807 210,594 210,266 Temporary worker population (thous) 132,433 132,013 131,485 130,908 130,361 129,894 129,529 129,221 128,943 128,639 128,266 Annual regular worker cost per person ($thous) 26.2 27.1 28.6 30.3 32.1 34.1 35.6 37.2 38.8 40.5 42.3 Annual temporary worker cost per person ($thous) 13.1 13.5 14.3 15.1 16.0 17.0 17.8 18.6 19.4 20.3 21.1 Total worker cost ($bn) 7,329 7,567 7,967 8,409 8,888 9,419 9,824 10,240 10,683 11,138 11,597

Work force demand (thous) 133,095 133,620 134,047 134,439 134,836 135,262 135,731 136,250 136,823 137,457 138,157 Worker's productivity growth, % 5.3% 5.0% 4.7% 4.4% 4.1% 3.9% 3.7% 3.5% 3.3% 3.1% Total Population 2,294,289 2,303,647 2,312,542 2,320,895 2,328,621 2,335,659 2,341,973 2,347,567 2,352,484 2,356,791 2,360,544 GDP per capita 14,719 15,145 15,570 16,000 16,436 16,879 17,328 17,783 18,242 18,704 19,171 GDP growth, % 3.3% 3.2% 3.1% 3.1% 3.0% 2.9% 2.9% 2.8% 2.7% 2.7%

Needed additional temporary worker (thous) 662 1,607 2,562 3,531 4,475 5,368 6,202 7,029 7,880 8,817 9,891 Co-bots effectivity (conversion into manpower) 1.6 1.7 1.7 1.8 1.9 2.0 2.1 2.1 2.2 2.3 2.4 Co-bots install base demand unit (thous) 16 115 251 421 615 827 1,046 1,267 1,479 1,712 1,977 Substitutable ratio by Co-bots 4% 12% 17% 22% 26% 30% 35% 39% 42% 45% 48% Substitutable ratio by Industrial robots 96% 88% 83% 78% 74% 70% 65% 61% 58% 55% 52% 10987654321 Co-bots real install base unit (thous) 5 39 102 193 307 440 590 746 899 1,060 1,245 Penetration rate 29% 34% 41% 46% 50% 53% 56% 59% 61% 62% 63% Co-bots density (unit/10,000 workers) 0 2 5 9 14 21 28 35 43 50 59 YoY growth, % 763% 161% 89% 60% 44% 34% 27% 21% 18% 18%

Co-bots annual cost per unit ($thous) 17.9 17.5 17.2 16.8 16.5 16.1 15.8 15.5 15.2 14.9 14.6 YoY growth, % -2.0% -2.0% -2.0% -2.0% -2.0% -2.0% -2.0% -2.0% -2.0% -2.0% Co-bots work force 7 66 178 349 580 867 1,211 1,596 2,004 2,463 3,016 Initial Co-bots cost ($mn) 81 606 1,065 1,484 1,819 2,050 2,228 2,231 2,087 2,129 2,381 Additional Industrial robots cost ($mn) 67 376 450 494 489 448 380 313 250 230 222

Temporary worker + Co-bots work force 132,590 133,090 133,745 134,557 135,497 136,521 137,579 138,602 139,549 140,499 141,479 YoY growth, % 0.4% 0.5% 0.6% 0.7% 0.8% 0.8% 0.7% 0.7% 0.7% 0.7%

Total cost ($bn) 7,329 7,568 7,969 8,411 8,890 9,422 9,826 10,242 10,686 11,140 11,600 YoY growth, % 3.3% 5.3% 5.5% 5.7% 6.0% 4.3% 4.2% 4.3% 4.3% 4.1%

New Co-bots sales ($mn) 81 606 1,065 1,484 1,819 2,050 2,228 2,231 2,087 2,129 2,381 Co-bots replacement sales ($mn) 0 20 168 429 793 1,239 1,741 2,287 2,834 3,345 3,867 Co-bots replacement unit (thous) 0 1 10 26 48 77 110 148 187 225 265 Replacement rate 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% Total Co-bots sales ($mn) 81 626 1,234 1,914 2,612 3,289 3,969 4,518 4,921 5,474 6,247

Source: Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 24 April 13, 2016 Profiles in Innovation

Scott Fetzer Electrical Group has optimised production costs by 20% by using Universal Robot’s cobots

The challenge: Tennessee-based Scott Fetzer Electrical Group (SFEG) manufactures a wide range of products in low volumes, meaning that manufacturing tasks are constantly changing depending on demand. SFEG was under pressure to remain competitive globally as the Asian electronics market expanded and wanted to avoid having to upgrade its entire existing asset base. Given the high-mix, low-volume nature of order intake, traditional industrial robots were not an economically viable option as lines would be unused for long periods of time.

The solution: Universal Robots’ model UR5 was adapted to work on a variety of tasks along with human operators, when more capacity is needed. According to IFR, for example, one day the robot could be bedding metal sheets, the next it could be performing pick and place tasks. The robot does not require a safety guarding like traditional industrial robots; the robot arm will automatically stop operating if it encounters obstacles. To build on the collaborative potential of the new robots, SFEG has put them on top of AGVs and now has a fleet of mobile UR robots deployed throughout the sheet metal department, integrating them into the entire production cycle, from cutting the initial blank on the blanking press to forming, folding and final assembly of the electrical components. Multiple robots are also being connected to work together on complementary tasks, such as moving parts between workstations. Live testing of final products was also automated. The cobots can turn the product on/off, run it for a couple of seconds and repeat that task for a full cycle of testing (up to 400 hours). Furthermore, the robot collects live data from all tests across several important variables. Preparing the robots for their tasks is also a much simpler process than for traditional industrial robots: this can be done through a simple user-friendly screen or by simply grabbing the robot arm and performing the desired task. SFEG says it took 30%-50% less time than with traditional robot implementations.

The result: According to the company, before it had the cobots on the transformer line, an operator could make on average 10 parts per hour; collaboration with the robots increased this productivity by 20%. SFEG says it has won back market share against Chinese competitors and brought back to the US some of its Chinese-sourced manufacturing as a consequence. The company says the payback for UR robots was 12-14 months.

Glory (Japanese machinery maker) deploys cobots on cash register assembly lines

The challenge: Obtaining production line workers for factories to manufacture cash registers became a problem for Glory in various countries because of declining/ageing populations.

The solution: The company introduced Kawada Technologies’ humanoid robot, Nextage, on assembly lines in 2012. Cobots were given nicknames that matched local currencies (Yen-chan, Dollar-kun, Euro-kun) to help workers feel comfortable with their new colleagues. Nextage is highly adept at simple picking work, like many cobots, but it also has the dexterity to tighten screws and mount components. We view this as significant in that it shows the potential for higher-value-added operations to be entrusted to cobots.

Exhibit 27: Universal Robots cobot alongside a worker… Exhibit 28: …and in a production line

Source: Teradyne. Source: Teradyne.

Goldman Sachs Global Investment Research 25 April 13, 2016 Profiles in Innovation

PHYSICAL MANUFACTURING

Additive manufacturing • Lower raw material consumption, quicker time-to-market and allows mass-customisation US$25-30 bn (2020E base case) • Early commercialisation stage • Savings example: Airbus production cost of satellites reduced by 20% by using in-house 3D printer as opposed to outsourcing production

Additive manufacturing or 3D printing is the process of making a physical object from a three-dimensional digital model, typically by laying down many successive thin layers of a material. The technology facilitates several aspects of the evolution of the global economy, and more specifically manufacturing, via: (1) reduced raw material consumption; (2) a quicker design to production process; and (3) mass customisation. Within 3D printing, printers and materials are the biggest opportunities, and we believe the industry could move towards a razor/razor blade model as it matures. This means that the installed base could become the driver of a more recurring and profitable opportunity, in the form of materials sales and technology. Competitive We highlight two public companies, 3D Systems and Stratasys, but also the leading landscape private company EOS. Other players include Renishaw and Hewlett-Packard. Han’s Laser offers indirect exposure.

The challenges Speed, energy consumption and further technological progress: (1) slow print speed is an impediment to high-volume production; (2) lack of material science advancement blocks the use in several applications (e.g. metals); (3) high energy costs can affect economics; (4) software development is in the early stages and needs to evolve in order to facilitate the design of more complex products; and (5) there are safety/environmental concerns around melting thermoplastic, which is one of the most used materials currently.

Applications and Similar to other manufacturing innovations such as robotics and vision systems, auto industries exposed companies were the early adopters of 3D printers and remain the biggest users. The key application for 3D printing is rapid prototyping of new models. Going forward, we believe autos will likely remain the biggest application, but see other end markets playing a major role in the growth of the industry, namely aerospace and consumer goods.

GE is leading the way in terms of commercial applications within aerospace. Specifically, GE is deploying 3D printed fuel nozzles and sensors for the GEnx jet engine. These components serve key purposes, allowing sensors to be mounted on parts that were previously difficult to monitor, the ability to work in hazardous environments (2000 Fahrenheit) and the use of materials not traditionally associated with manufacturing. Furthermore, companies like Siemens are using it as the next step after PLM simulation.

Sizing the revenue Additive manufacturing has a potential addressable market of US$25-30 bn by 2020E. opportunity Per IDC, the global 3D printing market is worth c.US$11 bn currently, and is expected to grow to c.US$27 bn by 2019, implying a 25% CAGR. We take a more conservative approach in our base case, and assume a 20% CAGR through 2015-20 and a terminal rate of 5% by 2025. This yields a market size of US$40 bn by 2025, implying a CAGR of 14%.

Exhibit 29: Estimating the potential of the 3D printing market US$ bn

Additive manufacturing 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E

3D printing potential revenue $11bn $13bn $16bn $19bn $23bn $27bn $30bn $33bn $36bn $38bn $40bn % yoy 20% 20% 20% 20% 20% 10% 10% 10% 5% 5%

2015 - 2025 CAGR: 14%

Source: Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 26 April 13, 2016 Profiles in Innovation

Simplification and weight savings enabled with 3D printing of commercial aviation parts

The challenge: General Electric requires nozzles that hold the T25 sensor which monitors GE’s Leap jet engine. Previously, the nozzles, like those used in the Leap engine, were machined from 20 separate parts, resulting in problems with ice accumulation and welding and joint strength. With 19 fuel nozzles per engine and high development costs, GE needed to streamline the manufacturing of the nozzles while satisfying strict aviation regulations.

The solution: With additive manufacturing technology, the newly certified part is made in one solid piece, reducing manufacturing and development time (approximately one year) while improving part performance.

“Once we found a workable solution, it went straight to production. This technology is a breakthrough.” – Jonathan Clarke, Program manager for the project.

Exhibit 30: 3D printed component used in GE’s Leap jet engines

Source: General Electric.

Goldman Sachs Global Investment Research 27 April 13, 2016 Profiles in Innovation

IN-FACTORY LOGISTICS

Automated Guided • Autonomous material handling equipment eliminating labour costs • Early development stage Vehicles • Savings example: Harley Davidson reduced costs 7% by introducing AGVs and creating a more efficient production facility c.US$2 bn (2020E base case)

Within manufacturing, automated guided vehicles (AGVs) are used in the automation of material handling, transporting objects through factories and distribution centres. As the technology advances, they should be increasingly connected to the broader IoT network, allowing for just-in-time delivery of components. As a result, the technology should help reduce labour costs and allow for more flexible production that does not need to follow a fixed conveyance line. Egemin, the AGV brand of Kion, estimates the payback time of an AGV system is 2.75 years. We see AGVs as an important development in an increasingly connected and automated logistics system, potentially leading to much lower levels of work-in-progress and inventories.

Competitive We highlight four companies: Siasun, Kion, Dematic and Kuka. Other companies landscape pursuing this technology include Toyota, Jungheinrich, Amazon (formerly Kiva Systems) and JBT Corporation. Private companies such as Savant Automation, Clearpath and Seegrid Vision are also making inroads in this market.

The challenges Infrastructure costs, potential new entrants and safety. This technology must first overcome a number of challenges: (1) widespread adoption may require large infrastructure spend on factory layouts to ensure safe use, multiplying the cost; (2) it opens up to competition similar technology used in autonomous cars; (3) similar to autonomous cars, it needs to overcome safety concerns in a likely highly hazardous environment; and (4) the technology still needs to advance and costs need to fall.

The markets disrupted This technology will disrupt other material handling equipment (such as forklifts). For some suppliers, it may cannibalise their primary offering, as evidenced by large investments from some of the leading players (e.g. Kion acquiring Egemin). Sizing the revenue The market for AGVs in industrial use could be worth c.US$2 bn by 2020 and almost opportunity US$4 bn by 2025. We assume the current installed base of manned material handling equipment as the potential market for this equipment (Class 2 and 3 industrial trucks). Using industry data, we estimate the proportion of this equipment used in manufacturing and logistics (57%), and forecast global demand to grow at long-run GDP (assumed at 3.0%). Using a variety of other industry sources and company data, we estimate the current AGV installed base is less than 2%; we expect this to grow to c.15% by 2025. We expect pricing to hold up for the next couple of years, as the technology makes significant advances, and to begin deflating once there is broader adoption.

Exhibit 31: Estimating the potential of the AGV market within manufacturing US$ mn

Automated Guided Vehicles 2014E 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E

Comparable material handling equipment manufacturing 342,000 352,260 362,828 373,713 384,924 396,472 408,366 420,617 433,235 446,232 459,619 473,408 YoY growth, % 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0%

Penetration of AGVs 1.2% 1.6% 2.1% 2.9% 4.2% 6.2% 7.8% 9.8% 11.4% 13.3% 14.2% 15.2%

AGV units sold 4,104 5,540 7,757 10,859 16,289 24,433 31,763 41,292 49,550 59,461 65,407 71,947 YoY growth, % 35% 40% 40% 50% 50% 30% 30% 20% 20% 10% 10%

Average cost of AGV + installation costs ($'000s) $65 $65 $65 $65 $64 $62 $61 $60 $59 $58 $56 $55 YoY growth, % 0%0%0%-2%-2%-2%-2%-2%-2%-2%-2%

Total AGV sales $267m $360m $504m $706m $1,038m $1,525m $1,943m $2,476m $2,911m $3,424m $3,691m $3,979m YoY growth, % 20% 20% 14% 14% 8% 47% 27% 27% 18% 18% 8%

Source: Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 28 April 13, 2016 Profiles in Innovation

Harley Davidson: Automated guided vehicles

Harley Davidson took to transforming its York Vehicle Operations facility, which accounts for more than 60% of the company’s motorcycle production.

The challenge: The original factory, laid out on mass-production principles, was becoming inefficient and inflexible. Harley Davidson set as its objective the adoption of lean principles and reduced manufacturing floor complexity.

The solution: One key aspect of this transformation was removing the assembly line production chain and adopting a digital chain (where assembly parts move on flexible automated guide vehicles (AGVs) which are driven by planning needs and software). There was a significant reduction in inventory at hand levels, which currently stand at three hours compared with 8-10 days in the legacy systems, owing to a major cut in the planning cycle, moving from a 21-day fixed plan to only a six-hour time horizon.

Moreover, the low-value-add, repetitive works have been moved to robotic execution, increasing productivity. Incorporation of a real-time performance management system helps in monitoring the entire manufacturing floor via multiple connected devices. This transformation resulted in significant metric improvements for the company: it reduced costs by 7% and productivity rose 2.4%, resulting in a net margin improvement of 19%.

Exhibit 32: An example of an AGV supplied by Siasun Exhibit 33: An example of an Egemin AGV

Source: Siasun. Source: Wikimedia.

Goldman Sachs Global Investment Research 29 April 13, 2016 Profiles in Innovation

IN-FACTORY LOGISTICS

• Allow for identifying and tracking objects in complex and connected Radio Frequency networks Identification • Mature technology with new early applications in manufacturing • Savings example: Griva, a textile manufacturer, achieved a ROI of 30% c.US$2 bn (2020E base case) in 9m from using Alien Technologies’ RFIDs to monitor inventory

Radio frequency identification (RFID) technologies are tags that use radio waves to transfer data, to allow the identifying and tracking of objects. The technology has existed since at least the 1970s. Until now, it has been too expensive to be practical for many commercial applications within manufacturing. The technology solves many of the problems associated with bar codes, as radio waves travel through most non-metallic materials, and can be embedded in components. The tags can store data about each component manufactured and allow tracking of their real time locations throughout the production process (e.g. for work-in-progress, supplies and product inventory).

Active RFID tags use a battery and run an imbedded microchip, submitting signals to the reader. They can be scanned over long ranges but cost a dollar or more. Passive tags have no battery. Instead, they draw power from the reader. Manufacturing companies are focusing on passive ultra high frequency (UHF) tags, which cost less than 50 cents today in orders of one million or more. Their read range is typically less than 20 feet vs. 100 feet or more for active tags, but they are far less expensive than active tags and can be disposed of with the product packaging.

Competitive landscape We highlight four of the largest specialist manufacturers: Zebra Technologies, Impinj, Smartrac and Alien. We also note some larger companies that provide this technology or components: DHL, FedEx and Honeywell (through Intermec). Software providers (e.g. Oracle and SAP) provide integration solutions.

The challenges Cost deflation, compatibility and technological barriers. There are still a number of barriers to widespread adoption within manufacturing: (1) even at 50 cents for a UHF tag, costs need to come down substantially in order for it to be economically viable for widespread adoption – currently their use is limited to only high-value items; (2) the RFID technology must be compatible with the broader IoT system protocols and standards; and (3) the technology still needs to overcome some barriers, such as not being able to send signals through metallic materials.

Applications and RFID are a key enabling technology for connecting factories and managing inventory levels. industries exposed They are currently used in many industries around the world and on a variety of applications. Club Car, a maker of golf carts, uses RFID to improve efficiency on its production line. Paramount Farms, one of the world’s largest suppliers of pistachios, uses RFID to manage its harvest more efficiently. Much of the recent growth has come from, and is continuing to come from, uses in retail apparel, and the largest orders are placed by governments. Food & beverages and textiles are positioned well, owing to the low number of metallic components. Sizing the revenue We estimate the Factory of the Future will offer a c.US$2 bn opportunity for RFID opportunity providers by 2020. In 2014, around 7.5 bn total RFID units were sold, with between 3% and 6% of these sales made to manufacturing (mainly UHF tags). Using publically available data from RFID Journal and other industry sources, we are able to forecast trends and pricing. We estimate that the growth rate of manufacturing units sold will be over 20% pa for the next five years. Despite this high rate, this is lower than is expected for the retail industry (the largest end market for RFID technologies, which we expect to grow at a CAGR of over 30%). We hold the average number of readers per 1,000 tags fixed in our forecasts, owing to “reader collision” blocking signals and limiting the number of readers possible in a fixed space. Note that given the growth in different sectors in our base case, by 2020E manufacturing still represents only 4% of total RFID unit sales, and therefore the total market for RFID should be much greater.

Goldman Sachs Global Investment Research 30 April 13, 2016 Profiles in Innovation

Exhibit 34: Estimating the potential of the RFID market within manufacturing US$ mn

RFID 2014E 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E

RFID units sold (bn) 7.5 9.4 11.7 14.6 18.3 22.9 28.6 34.3 39.5 43.4 45.6 47.9 YoY growth, % 25.0% 25.0% 25.0% 25.0% 25.0% 25.0% 20.0% 15.0% 10.0% 5.0% 5.0%

% of which in manufacturing 4.5% 4.4% 4.3% 4.2% 4.1% 4.0% 3.9% 3.8% 3.7% 3.6% 3.5% 3.4%

RFID manufacturing units sold (mn) 337.5 413 504 615 751 916 1116 1305 1461 1563 1596 1628 YoY growth, % 22.2% 22.2% 22.1% 22.0% 22.0% 21.9% 16.9% 12.0% 7.0% 2.1% 2.0%

Average cost of UHF RFID ($) $0.50 $0.48 $0.45 $0.43 $0.41 $0.39 $0.37 $0.35 $0.33 $0.32 $0.30 $0.28 YoY growth, % -5.0% -5.0% -5.0% -5.0% -5.0% -5.0% -5.0% -5.0% -5.0% -5.0% -5.0%

Total manufacturing RFID tags sales $169m $196m $227m $264m $306m $354m $410m $456m $485m $493m $478m $463m YoY growth, % 16.1% 16.1% 16.0% 15.9% 15.9% 15.8% 11.1% 6.4% 1.7% -3.0% -3.1%

Average number of readers per 1000 tags 222222222222

Average cost of UHF reader $1,000 $950 $903 $857 $815 $774 $735 $698 $663 $630 $599 $569 YoY growth, % -5.0% -5.0% -5.0% -5.0% -5.0% -5.0% -5.0% -5.0% -5.0% -5.0% -5.0%

Total manufacturing RFID reader sales $675m $784m $910m $1,055m $1,223m $1,417m $1,640m $1,822m $1,938m $1,971m $1,911m $1,852m YoY growth, % 16.1% 16.1% 16.0% 15.9% 15.9% 15.8% 11.1% 6.4% 1.7% -3.0% -3.1%

Total manufacturing RFID sales $844m $980m $1,137m $1,319m $1,529m $1,771m $2,051m $2,278m $2,423m $2,463m $2,389m $2,315m 16.1% 16.1% 16.0% 15.9% 15.9% 15.8% 11.1% 6.4% 1.7% -3.0% -3.1%

Source: Goldman Sachs Global Investment Research.

Bosch RFID tags cut inventory time by 97% in China

The challenge: In the Bosch plant in the Chinese city of Suzhou, the yearly task of taking machine inventory used to be a major undertaking. Plant 1 has four manufacturing areas, each with up to 2,500 machines, test benches, and items of measuring equipment. For ABS manufacturing alone, the inventory process used to take up to a month in some cases. Sometimes associates printed out lists to help them manually record machine inventory.

The solution: Now, thanks to smart connectivity, inventory taking takes just four hours. All the machines and equipment items have been fitted with RFID transponders. This allows objects to be identified without physical contact. Now, associates push RFID trolleys fitted with a laptop and antennas through the manufacturing shop (in the future AGVs could also do this too). As they move along, the trolleys use RFID technology to automatically identify machines and devices. This cuts the time needed for inventory taking by 97%, or 440 man hours.

Stanley Black & Decker’s Mexico plant

The challenge: Stanley Black & Decker wanted to improve the overall operating efficiency of one of its largest plants in Reynosa, Mexico.

The solution: It implemented real-time location systems (RTLS) on its production lines. RTLS include small and easily deployed RFID tags that provide real-time location and status, thereby allowing better work efficiency, better supervision of inventory, reduced working capital and improved production line throughput. Stanley Black & Decker had a 15% increase in revenue per employee in 2014 and improved equipment effectiveness, resulting in a 300 bp operating profit margin improvement.

Goldman Sachs Global Investment Research 31 April 13, 2016 Profiles in Innovation

Other key technologies (ordered alphabetically)

We briefly highlight five other technologies and key enablers that could have a dramatic impact on the future of manufacturing:

Augmented reality vision Augmented reality vision is an upcoming technology that superimposes a computer- generated image on a user’s view of the real world. This technology could allow engineers and workers a view of 3-dimensional designs and let them receive visual instruction to complete their tasks, increasing productivity and reducing the likelihood of costly errors. This could disrupt the market for 2-dimensional computer screens used to instruct engineers currently.

In Profiles in Innovation Vol. 1: Virtual & Augmented Reality, January 13, 2016, GS analysts estimated that, despite no adoption until 2017, and including virtual reality devices, this market could be worth US$1.5 bn in 2020 and over US$4.5 bn in 2025. This technology is also led by tech giants such as Facebook, Samsung, Google, and Microsoft.

Demand response Demand response provides an opportunity for factories to play a significant role in their own consumption from the electric grid by reducing or shifting their electricity use during peak periods in response to time-based rates or other forms of financial incentives. Given their large energy use, factories often have the most to gain from any commercial and industrial customer by optimising their energy consumption and helping reduce costs. The latest battery technology can help maximise this potential. Key to this technology is the use of energy storage, which in The Great Battery Race, October 18, 2015, GS analysts estimated was a TAM of US$100-150 bn, of which the demand response is worth US$45-71 bn. Key players include Tesla, Saft Batteries and Schneider Electric.

Machine learning Machine learning (or artificial intelligence) is the ability of machines to process data into information and derive knowledge from that information to act independently or augment human decision making. The key difference between machine learning and a smart piece of code or smart connected devices is it is capable of self-learning/improving.

Machine learning has been introduced at a basic level in collaborative robots, and is touted as a key enabler in optimising the IoT. Within IoT ecosystems, machines are able to communicate with each other and adjust/work independently of human interaction. Key players are the tech giants (Google, Microsoft, IBM) and GE, although there are numerous exciting start-ups in this space across a variety of different applications.

Machine vision Machine vision is the technology and methods used to provide imaging-based automatic inspection, gauging, counting and analysis at high speeds, reliability and with greater precision (exceeding the capabilities of the human eye).

This technology, which effectively consists of highly advanced sensors, has the potential to dramatically increase the capability of other technologies such as collaborative robots, AGVs, and other automated production processes. Key players include: Cognex, Omron, Keyence and Google (through Industrial Perception).

Nanotechnology Nanotechnology is engineering on an atomic scale. The analysis or manipulation of atoms and molecules is key to addressing the challenge of extracting higher output from less resource, and is becoming increasingly important in manufacturing. Nano materials are already used in consumer goods such as make-up, and industrial products such as surface coatings. In addition, the semiconductor roadmap provides a dramatic example of this in the relentless miniaturisation and increasing power of electronic devices.

Goldman Sachs Global Investment Research 32 April 13, 2016 Profiles in Innovation

As the limitations of existing technologies become more binding, we believe the importance of this technology will increase, in turn driving an increasing number of nanotechnology applications (e.g. purifying silicon semiconductors). Nanotechnology accounts for a small but growing share of public and corporate research budgets; Oxford Instruments is a public company with larger exposure.

Autodesk is using augmented reality to enhance computer aided design (CAD)

Our US technology analysts spoke to Autodesk to discuss the augmented reality used in manufacturing to visualise 3D models in a real-world context, avoiding costly errors and improving collaboration.

The challenge: While CAD software has done much to improve the design process it still has its limits. For example, CAD is still generally constrained by the need to work on a 2D monitor. This limits the ability to view full-sized objects in a real-world context, and hinders true collaboration when many people are working on the same design.

The Autodesk solution: Autodesk is a leading design software company with US$2.5 bn in annual revenues and with customers primarily in the architecture, engineering, construction and manufacturing verticals. Our analysts spoke with Autodesk’s emerging technology division, which is working with Microsoft HoloLens to incorporate AR into CAD. Autodesk’s VRED 3D visualisation software product is currently used by auto makers to project things such as doors and colours onto clay car models.

Wide-ranging applications: Autodesk envisions AR improving the design of items as small as videogame controllers or as large as buildings, with the view that having the true 1-to-1 scale that’s not possible in 2D can go a long way towards avoiding errors. Autodesk also sees value in looking at a 3D model and having metadata information behind the CAD at your fingertips. Finally, if two people are collaborating on a project, they are better able to work through issues as they go as opposed to working on separate desktops and realising there are issues when they come together. Given the wide range of CAD use cases, we see potential for AR to impact Autodesk’s base of 5 mn customers worldwide.

Potential bottlenecks: The computing power of AR systems needs to improve to run large CAD files.

Exhibit 35: Autodesk is working with Microsoft

Source: Microsoft.

Goldman Sachs Global Investment Research 33 April 13, 2016 Profiles in Innovation

Industries: Top-down: >US$500 bn addressable market

 Higher labour costs and increased global competition and putting pressure on manufacturers to automate and making time to market increase in importance as a differentiator

 We see the greatest opportunity from a shift to best-in-class

manufacturing for: o Electronics – Up to US$120 bn o Machinery – Up to US$110 bn o Food & Beverages – Up to US$100 bn

Goldman Sachs Global Investment Research 34 April 13, 2016 Profiles in Innovation

Total addressable market could be worth more than US$500 bn

We believe the total addressable market for equipment manufacturers supplying physical automation and software for full connectivity and optimisation of factories could be worth more than US$500 bn, or 10% of fixed investment, in cost savings. We assess this market size on a top-down basis across the relevant industries, based on current levels of automation and digitisation and potential savings achieved in existing technology pilots. This assessment provides a higher value than our bottom- up analysis of the six core technologies we presented earlier, as: (1) our top-down analysis includes other technologies (such as those that are more mature, but where penetration can still be increased or efficiency improved); (2) it aggregates synergies from using more than one of the six technologies profiled; (3) we use a savings approach, which will not necessarily be realised as revenues by equipment providers as some of the value identified will likely be shared with customers rather than manufacturers (especially for technologies where prices deflate quickly).

Approach to our top-down TAM assessment We believe the size of the total addressable market for all emerging technologies ensuring optimal automation and digitisation of future manufacturing is capped by the potential size of the savings achieved by using those technologies. This would be a best-case scenario, in which the industries adopting these technologies do not retain or pass through to their customers any achieved savings (the latter is what usually happens).

For our top-down assessment, we look at each industry’s level of automation and digitisation vs. the best-in-class example of the automotive industry. We expect manufacturing to evolve differently for each global industry. Currently, each is at a different stage of automation and sophistication owing to the competitive dynamics and market pressures it faces. For industries such as automotive, where factories are already highly sophisticated, the incremental value-add of new technologies will be quite low.

We then collect numerous examples of full automation, digitisation and optimisation achieved in pilots in the eight industries we analyse. Crossing those two data points, we estimate the potential savings achieved, which we assume equates to the maximum possible size of the addressable market of the technologies that allow for those gains.

In more detail, our top-down TAM assessment is based on the following:

 Assessing the relevant pool of fixed investment. Manufacturing accounts for around 25% of global gross fixed capital formation (or fixed asset investment). Using data from the World Bank, OECD and other national accounts, we are able to estimate the proportion of fixed asset investment by each manufacturing industry.

 Assessing the level of existing automation and digitisation. We capture the varying degrees of automation in each industry by looking at the current penetration of robots (in Japan, Germany, the UK, France and Italy) and software. This is important in assessing the maximum potential for other manufacturing industries if they were to catch up with the more advanced ones, most notably automotive.

 Assessing the level of savings enabled by full automation/digitisation. Using Goldman Sachs research, case studies and third-party sources, we estimate a range of gross savings that each industry could generate by introducing and implementing the latest technologies. For some industries such as autos, which have already automated significantly, the extra savings are quite limited; for others, the potential is much greater. We also apply a relevance factor for each technology.

 Calculating the maximum revenue pool potential for equipment manufacturers. Given the proportion of investment available, the potential savings for each industry and the degree of automation already in place, we estimate the maximum potential savings and revenue pool available.

Goldman Sachs Global Investment Research 35 April 13, 2016 Profiles in Innovation

 Qualifying speed of adoption. It is unlikely that each industry will automate and innovate to the extent the automotive industry has. Looking at the competitive structure, age of assets and the margin pressure that industries are facing, we also incorporate a metric that aims to capture the likelihood of this maximum revenue pool per industry being realised. Unsurprisingly, the automotive industry is highly likely to adopt new technologies, while the food products sector is less so, for reasons highlighted further on.

Exhibit 36: We look at the key industries that drive fixed Exhibit 37: …assessing their software penetration… investment … Software as % private non-resi fixed asset investment for Global fixed investment broken down by industry various industries in the United States (2015)

Others / Not 18% Textiles, leather & classified, 10% wearing apparel, 16% 1% 14% Global Wood and paper 12% products, 4% manufacturing Automotive, 19% GFCF: ~$4.75tn 10%

Metals, 8% 8% 6% 4% Food products, 2% beverages & Chemicals and tobacco, 8% plastics, 21% 0% Softwareof fixed% as asset investment Metals Total plastics Machinery classified Electronics, 20% Electronics tobacco products

Machinery, 10% Not / Others Chemicals & Chemicals beverages & Manufacturing Wood & paper Food products, transport equip. Motor vehicles & vehicles Motor wearing apparel wearing Textiles, leather&

Source: World Bank, OECD, China Statistical Yearbook, Goldman Sachs Global Source: US Bureau of Economic Analysis, Goldman Sachs Global Investment Investment Research. Research.

Exhibit 38: …as well as level of automation… Exhibit 39: …and urgency of need for optimisation Robot density per 1,000 employees Net margins in 2010-14 and 2015-18E for selected manufacturing industries

12.0% Automotive 80 - 120 Net margins (2010-14) Electrical/electronics 20 - 40 10.0% Net margins (2015-18E) Plastic and chemical products (inc. pharma) 15 - 35 8.0% Metals 5 - 15

Machinery 5 - 15 6.0% Food products beverages; < 8 Tobacco products 4.0% Wood & paper products < 5 Textiles, leather, wearing 2.0% < 1 apparel Total manufacturing 15 -25 0.0% Electronics Food Machinery Chemicals Automotive Wood and Metals 0 20406080100120 products, and plastics paper beverages & products Robots per 1,000 employees tobacco

Source: Goldman Sachs Global Investment Research. Source: Company data, Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 36 April 13, 2016 Profiles in Innovation

Exhibit 40: We used 60+ examples of savings captured by introducing FoF technologies in their current state

Company Fanuc FDM Flextronics Sub‐Zero Maserati Continental Stanley Black & Decker Large savings & faster time‐ Business Outcome Lower downtime Reduce energy Faster time‐to‐market Shorten time‐to‐market Lower Inventory Reduce defects to‐market Technology introduced IoT 3D printing IoT IoT PLM software IoT IoT Use of 3D printing to Machine level Use PLM software to Monitoring & predictive Collaboration enhanced Real time supply chain Big data enabled quality Description manufacturer component consumption & demand design its complex maintenance across platforms monitoring & interaction checks used in drones response products

Old state 11% unplanned downtime 20 days outsourcing $8.4mn/factory 15 months cycle 30 months time‐to‐market 14x inventory turnover 4.9% defect rate

2 days with a ROI of New state 5.8% unplanned downtime $6.9mn/factory 11 months cycle 16 months time‐to‐market 19x inventory turnover 2.5% defect rate $12,000 Improvement 47.8% 10x faster & ROI in 9m 17.5% 23.0% 47.0% 34.8% 48.9% Savings of $40mn from Savings of $800,000 cost Savings of $1mn per 15% reduction in cycle Helped Maserati produce Reduced certain Result DPM reduced by 16% single customer avoidance over 3 years factory time 3x more cars component costs by 20%

Source: Company data, Goldman Sachs Global Investment Research.

We also perform a scenario analysis to calculate the savings based on the assumption that using labour-reducing technologies (AGVs and cobots), labour-intensive industries (such as textiles, wood and food products) are able to halve their gap with the automotive industry. Using data from five of the ten largest manufacturing countries – Japan, Italy, the UK, France and Germany – we follow five steps: (1) calculate labour intensity (number of employees/output); (2) calculate the reduction in labour needed to halve the difference with the automotive industry; (3) assign high (US$30,000) and low (US$15,000) wages; (4) multiply wages by absolute labour reduction to calculate displaced savings; and (5) divide by output to find the margin impact.

Exhibit 41: Our scenario analysis suggests significant savings potential for textiles and wood products Analysis of hypothetical scenario in which industries halve their labour intensity (no. employees/output) difference with autos

Savings from labor displacement No. Labour Labour Wages Savings in Manufacturing industry Output ($m) % of output Employees intensity reduction (high/low) wages($m) Food products beverages; Tobacco products 3,773,158 273,462 14 20% Low 14,827 5% Textiles, leather, wearing apparel 1,278,993 52,649 24 33% Low 8,380 16% Wood & paper products 2,033,371 89,648 23 32% Low 12,825 14% Plastic and chemical products (inc. pharma) 3,276,175 401,736 8 0% Low 0 0% Metal (exc. Machinery) 3,779,939 290,443 13 18% High 20,209 7% Machinery 3,286,976 257,301 13 17% High 16,978 7% Electrical/electronics 2,953,577 268,034 11 12% High 10,629 4% Automotive 2,548,188 304,233 8 0% High 0 0% Others / Not classified 3,168,377 210,498 15 22% Low 14,053 7% Total manufacturing 26,098,754 2,148,004 12 16% Medium 101,344 5%

Source: BEA, OECD, World Bank, Goldman Sachs Global Investment Research.

Note: These estimates are not indicative of the revenue that we believe will be realised by equipment providers, but represent a maximum cap to those revenues. These estimates are also not fully risk- or probability-adjusted. In most case we use gross cost savings (although we attempt to net the costs of introducing new technologies where we can). While the price of key technologies may have declined significantly in the past few years, we do not fully incorporate the cost of restructuring the fixed infrastructure to accommodate this technology. As a consequence, our estimates are not a full representation of the costs that may be incurred. Industry level data will still vary significantly among countries; see Where will it be built? Page 49.

Goldman Sachs Global Investment Research 37

April 13, 2016April 13,

Goldman Sachs Global Investment Research Exhibit 42: Top-down analysis of bringing each industry’s manufacturing to best-in-class Industry-by-industry potential addressable market in US$ bn

Maximum Industries Current size of industry1 Current degree of automation 2 Potential savings 6 opportunity size % of global manuf. Robot intensity Robot Software % of Software Manufacturing sub- Automation Savings from FoF fixed asset Amount ($bn) per 1,000 penetration fixed asset penetration ($bn) sectors index technologies investment employees 3 index investment4 index 5 Automotive 19% 895 97 90% 9% 63% 76% 6 - 8% 50 - 70

Chemicals and 21% 990 19 19% 3% 17% 18% 5 - 7% 50 - 70 plastics

Electronics 20% 965 33 34% 17% 90% 62% 9 - 13% 90 - 120

Machinery 10% 485 8 9% 9% 54% 32% 17 - 23% 90 - 110

Food products, 8% 365 5 6% 5% 30% 18% 20 - 27% 70 -100 beverages & tobacco

Metals 8% 360 10 11% 5% 30% 21% 16 - 21% 60 - 70

Wood and paper 4% 200 2 2% 8% 47% 25% 18 - 25% 40 - 50 products

Textiles, leather & 1% 50 0 0% 8% 49% 25% 20 - 27% c.10 wearing apparel

Others / Not classified 9% 440 13 14% 7% 42% 28% 8 - 10% 40 - 50

Total Manufacturing 100% 4750 21 21% 8% 47% 34% 10 - 15% 500 - 650

Industries Pressure to automate Current technology penetration 7 Relevance factor

Manufacturing sub- Net margins Net margins 30yr average Current age (100 = very IoT PaaS PLM IoT PaaS PLM ∆ net margins % difference Cobots 3D printing AGVs RFIDs Cobots 3D printing AGVs RFIDs sectors (2010-14) 6 (2015-18E) age 7 (2015) likely) software software software software Automotive 4.9% 5.4% 0.5% 6.1 6.5 6.3% 85 ◕◕◔◑◔◑ 1.00 1.00 1.00 1.00 1.00 1.00

Chemicals and 7.0% 7.5% 0.5% 8.0 7.9 -1.4% 72 0.25 1.00 0.00 0.00 0.00 0.00 plastics ◔◕○○○○ Electronics 10.2% 10.0% -0.3% 7.5 9.5 27.2% 80 ◑◑◔◔◔◔ 1.00 1.00 1.00 1.00 1.00 1.00 Machinery 7.9% 6.3% -1.6% 9.3 10.5 13.1% 86 ◔◑◔◑◔◑ 1.00 1.00 1.00 1.00 1.00 1.00

Food products, 9.3% 9.9% 0.6% 8.5 8.6 1.4% 68 1.00 1.00 0.75 0.50 1.00 1.00 beverages & tobacco ◔◔○○◔◑ Metals 2.0% 2.7% 0.7% 11.8 11.4 -3.7% 72 ◔◔○○◔○ 0.75 1.00 0.00 0.50 0.75 0.50

Wood and paper 4.8% 7.0% 2.2% 8.0 9.2 13.9% 74 1.00 0.75 0.50 0.25 1.00 0.75 products ◕○○○◔○

Textiles, leather & n.a. n.a. n.a. 9.2 12.2 33.0% n.a. 1.00 0.75 1.00 0.50 1.00 1.00 wearing apparel ○◔○◔◔◔

Notes: 1 GFCF data from the World Bank, broken up using data from the OECD and China Statistical Yearbook 5 Automotive software penetration adjusted up according to analyst discretion 2 6 Robot penetration indexed against automotive; software penetration indexed against electronics Net margins of global GS coverage in each industry, note not entire industry and some companies in the sample have exposures to other industries Profiles in Innovation 3 Weighted average of density in Germany, Japan, France, Italy and UK 7 Average age data is for the US only 4 Software as a % of investments in private non-residential fixed assets in the US; automotive includes other transport equipment 8 Qualitative assessment of penetration at analyst's descretion (actual penetration may be less or greater) 38 Source: Goldman Sachs Global Investment Research.

April 13, 2016 Profiles in Innovation

A shrinking addressable market over time We view the total addressable market as the upper limit of customer budgets (as typically, capex/sales falls over time); accordingly, incremental value for capital goods providers shrinks over time as efficiency-driven innovation progresses and technology prices deflate.

Focusing on the total addressable market and potential business opportunities for capital goods companies, we look at the relationship between IT-related manufacturers and their equipment providers as an example. The life-cycle of IT-related products is short, and we believe the trend in this industry can be viewed as a proxy for other industries.

While markets for IT products such as flat panel displays (FPD) and semiconductors continue to expand, the scale of the markets for equipment used to produce these goods (measured by revenue as a percentage of total end-market capex) has declined to below the historical peak (Exhibit 43). Of interest here, in our view, is that customers such as FPD/semiconductor makers have tended to set investment amounts (or capex-to-sales ratios) for the purchase of capital goods at extremely high levels during the germination period for finished products (more than 20% of sales in some cases). Once this period ends, however, capex tends to enter a downward trajectory as the finished product market expands and matures.

Exhibit 43: Customer budgets have shrunk over the period Trends in FDP industry sales; % of FPD equipment sales vs. industry sales

140,000 30%

120,000 25%

100,000 20% 80,000 15%

($mn) 60,000 10% 40,000

20,000 5%

0 0% 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 FPD Industry Revenue (lhs) Equipment Intensity (%, rhs)

Source: Company data, Goldman Sachs Global Investment Research.

Value-add for the finished products of these manufacturers (customers of capital goods companies) expands as efficiency-driven innovations evolve through development of new technologies, improved yields, economies of scale, and enhanced technical skills among workers. However, selling prices for these products, particularly in IT-related B2C line-ups, nearly always decline, even with substantial evolution in embedded technologies. The performance of today’s FPD/semiconductor products is overwhelmingly superior to that of products from 10-20 years ago, and yet the value-added generated by investment in production equipment relative to total finished product is in decline.

FPD/semiconductor manufacturers are entering a mature stage in their industry life-cycle. They are moving towards near full automation in order to remove labour from the production process. However, for capital goods companies, it is apparent that: (1) customers will continue curbing budgets for capital goods beyond what is necessary unless there is ongoing development of new disruptive technologies; and (2) there is likely to be a shakeout of weaker players as capital goods makers compete for these limited capex budgets. Key for capital goods makers is providing optimal value-added by helping simplify processes to achieve the highest labour cost savings and other efficiencies to which they can tie the value of their products.

Goldman Sachs Global Investment Research 39 April 13, 2016 Profiles in Innovation

In the smile curve shown in Exhibit 44, which illustrates the general distribution of value- added along the supply chain in the manufacturing industry, production processes are located at the lowest point. In view of the downtrend of capex-to-sales throughout the life- cycle of each product, as mentioned above, it is not easy for capital goods producers to achieve an increase in payments from customers for value added. One solution to this problem is to strive to provide value added to capture investments allocated by customers to other areas. For instance, we think it is important to determine whether or not there are economic rationalisations to be achieved simply by taking fixed operating costs and turning them into depreciable costs.

Exhibit 44: Manufacturing is located at the lowest point Exhibit 45: Japanese customer budgets entered a of value add in the production process downward trajectory once manufacturing had matured Smile curve of value added in the production process 2-year moving average (capex + R&D)/sales for 30 of the largest manufacturing companies today

Higher 15% Japan 13% Europe US 11% Concept/R&D Sales/after service China

Value added Value Branding Marketing 9%

Design Distribution 7% Manufacturing Manufacturing creates Lower the least value 5% Production chain Time 3% 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015

Source: IEC. Source: Goldman Sachs Global Investment Research.

A typical example of where capital goods companies have been able to demand a greater share of customer budget was the substantial value-added achieved through introducing industrial robots in the autos industry. The cost of introducing robots is falling each year; we estimate standard upfront costs are now around US$100,000-200,000, including system integration costs. If the depreciation horizon (5-10 years) is taken into account, operating costs fall to a level comparable with labour costs (around US$20,000-30,000 per year). Depending on the size of the customer and production characteristics, the use of robots can have significant advantages over labour in view of their higher productivity and precision, as well as the potential to reduce trade union bargaining power and of course labour costs. In the 1980s, this rationale triggered a sharp increase in the uptake of robots; however, since this ‘revolution’, the capex-to-sales ratio for the industry has been in decline. For the electronics, food & beverage and machinery industries, we believe FoF technologies are capable of adding significant value and their share of customers’ budgets will respond accordingly.

If we scale this analysis up and look at the world’s largest manufacturers, using Japan as a case study, we can see how industrial development affects budgets. Since the 1980s, Europe and US customer budgets have been on a downward trajectory. In Japan, during the nascent years of its manufacturing industry, budgets rose and exceeded those of Europe and the US until the broader manufacturing industry matured and budgets also began heading into a downward trajectory. Currently, the (capex + R&D)-to-sales ratio for China’s largest manufacturing companies is still at 5% (Exhibit 45), highlighting where significant opportunities for equipment providers still exist.

Goldman Sachs Global Investment Research 40 April 13, 2016 Profiles in Innovation

Industry-by-industry deep dive

The future of the manufacturing ecosystem will differ markedly for each industry. Influencing factors include: (1) the current level of automation; (2) the market pressures industries are facing; (3) the nature of the products they produce; and (4) other factors such as labour intensity and asset age. Even within each industry, companies are likely to employ different levels of sophistication depending on the products that they manufacture and where the highest leverage in the cost structure stands. While autos has pioneered automation owing to thin margins, even there opportunities remain (e.g. for cross-plant integration). Within underpenetrated industries, we see the highest potential for automation in electronics, food & beverage and machinery.

Automotive – TAM up to US$70 bn We estimate that around 19% of global manufacturing fixed asset investment goes into the automotive industry, worth around US$900 bn. Regional hotspots include the US, Germany, Japan and China. The automotive industry has pioneered industrial automation from the early 1970s: it is by far the largest market for industrial robots, accounting for 43% of global robotics sales in 2014. Its companies have also been early adopters of the latest technologies highlighted in this report.

In our view, capital goods companies have limited scope to add incremental value in the autos industry given its already highly sophisticated production processes. However, we believe the competitiveness of the industry and the margin pressure that car manufacturers face will continue to place them at the forefront of adopting the latest technologies. Given that the market is dominated by a handful of OEMs, we expect automakers to continue to exert increased pricing pressure on their equipment providers, squeezing capital goods companies. We do not expect OEM capital goods budgets to get bigger, but rather to shrink, reflecting the relentless pressure to reduce costs (with our GS Capex Tracker pointing to a -0.8% capex CAGR for 2015-18E). When making the decision of where to manufacture products, the automotive industry is still highly influenced by tariffs, which are present in all major manufacturing locations. We believe the two key technologies for this industry given the current levels of penetration are increased usage of automatically guided vehicles and the introduction of collaborative robots.

Goldman Sachs Global Investment Research 41 April 13, 2016 Profiles in Innovation

BMW’s automation story

Stage 1: In the mid-1990s, BMW introduced its first large installations of industrial robots and leap-in automation. BMW transformed its production lines from car-body manufacturing to employing hundreds of robots and pushing the level of automation up to more than 80%.

This all came at the same time that substantial developments were made in the computer industry, which also had an impact on Programmable Logic Controller (PLC) technology. The basic automation concept was born to link a centralised PLC together with a number of robots. The robot was completely controlled by the PLC. With much larger installations, machine and human safety became an issue as the PLC capacity became an increasing constraint.

Stage 2: After 2000, Safety PLCs and Decentralised Periphery (DP) were introduced to eliminate this constraint and allow even larger installations with an automation level of up to 97%.

In parallel, the robot controller became more powerful and able to perform additional tasks such as safety-related operations. Software replaced hardware and allowed the robot to operate in a safer way with higher flexibility. At the same time, the mechanics got better, allowing a much higher degree of accuracy. New applications became possible, such as Remote Laser Welding, creating new possibilities for products.

Stage 3: Now, a typical installation at BMW involves between 800 and 1,000 industrial robots able to handle 10-750 kg payloads. Approximately 150 PLCs are connected to the robots and the Master Production Scheduling IT System, allowing one-piece production for up to six different car derivatives on one line. Data-Matrix and Bar-Codes are used to steer the manufacturing of the car body shell. The Master Production Scheduling is held in a Cloud and represents a comprehensive virtual overview of what is happening on the shop floor to guarantee that the customer’s car is delivered on time with the correct and individual specifications.

Source: International Federation of Robotics

Electronics – TAM up to US$120 bn We estimate that around 20% of global manufacturing fixed asset investment is into the electronics industry, worth over US$950 bn. In our opinion, this industry has the greatest opportunities despite its already relatively high levels of automation. If it were to fully introduce the disruptive technologies that we highlight, in particular within the final assembly line, then we believe maximum cost savings to the industry could equate to a TAM of around US$120 bn. This would be a very important contributor in an industry plagued by the challenge of constant product price deflation.

While not to the same extent as in the automotive industry, the manufacturing processes in electronics are already quite mature in terms of automation, and there are powerful customers within this industry. As a result, we predict a similar outcome for capital goods companies in electronics as in the automotive industry: we expect the absolute volume size to remain significant and the industry to be one of the early adopters of new technologies, but capital goods companies’ ability to claim some of the profit pool and the growth in customer budgets is likely be quite limited. While all FoF technologies have significant relevance in electronics, we believe the physical manufacturing technologies (cobots and 3D printing) and in-factory logistics (AGVs and RFID) have the greatest potential.

Goldman Sachs Global Investment Research 42 April 13, 2016 Profiles in Innovation

Machinery – TAM up to US$110 bn We estimate that around 10% of global manufacturing fixed asset investment is into the machinery industry, worth US$450-500 bn. Similar to the automotive and electronics industries, competition is based on innovation and quality. Machinery also consists of highly tradable goods, where manufacturers have access to global markets. However, volumes are substantially lower than in autos and electronics and machinery manufacturers are not able to take advantage of economies of scale in the same way that those two industries can – this is one of the major reasons why the level of automation is currently relatively low. As costs come down and capabilities improve, we expect to see much greater adoption rates. Starting from a low level of automation, we see machinery as having the potential to be one of the largest addressable markets. Similar to the automotive industry and electronics, all FoF technologies highlighted are relevant, but we see the most potential for the three with the lowest current penetration: IoT PaaS, AGVs and cobots.

Food products, beverages & tobacco – TAM up to US$100 bn We estimate that around 8% of global manufacturing fixed asset investment is into the food products, beverages & tobacco industry, worth US$350-400 bn. Food products, beverages & tobacco are manufactured in locations that simultaneously optimise proximity to raw materials and end customers. This is because the products must be fresh and appeal to local preferences. As a result, food manufacturing typically has quite low tradability, although some products such as powdered milk and frozen seafood are heavily exported. As in chemicals (and indeed other “The food industry as a whole is probably the next largest sector [for process industries), IoT cloud-based platforms for automation] but it is very diverse. holistic management of plant networks and the Robots are currently mainly being use of virtual modelling for plant design and risk used here in downstream mitigation during operations are likely to be the packaging, palletising and logistics more significant investments in the near term. and not very much in processes upstream of packaging. There is In our view, the FoF technologies AGVs, IoT great potential to use robots in PaaS, and PLM software offer the greatest future for handling and processing opportunity. PLM software is used by Coca Cola tasks, such as cutting, positioning,

to design bottles for better CO2 dispersion and by inspection. It is already being done MWV, a global packaging provider, to reduce the but much less than it could be.” development timeline from 18 to six months. Robotics expert from the International Federation of Robotics Corporation Metals – TAM up to US$70 bn We estimate around 8% of global manufacturing fixed asset investment is into the metals industry, worth more than US$350 bn. Metals plants are resource- and energy-intensive and their products are heavy and bulky, so the most important factors for success in those industries include easy access to raw materials, low-cost energy and inexpensive transportation. Similar to chemicals, the longevity of plants delays adoption and some of the technologies highlighted in this report are not appropriate for the industry (as it is characterised by flow type production vs. discrete batches). We believe PLM software is the most relevant for this industry, but there is also significant potential for AGVs and IoT PaaS.

Goldman Sachs Global Investment Research 43 April 13, 2016 Profiles in Innovation

Chemicals and plastics – TAM up to USS$70 bn We estimate that around 21% of global manufacturing fixed asset investment is into the chemical and plastics industry, worth US$950 bn-US$1 tn. Chemicals include bulky, commodity-type products with relatively low trade intensity, but also R&D-intensive pharmaceuticals and cosmetics that have high value density and tradability. The variety of products within this industry requires a number of different production environments, and the size of the opportunity for capital goods providers will vary accordingly.

Chemicals is the largest manufacturing industry in terms of fixed asset investment, and given its size it is a huge market for equipment providers. However, we believe that the chemicals industry will be especially slow to adopt the latest FoF technologies, if at all in some cases. First, a number of the technologies highlighted are directed towards use in discrete manufacturing industries rather than the production of continuous flow liquid or gas processes (e.g. cobots or AGVs). Chemicals scores relatively low on our “pressures to automate” metrics. And second, chemical refineries and plants are particularly risk averse, for two main reasons:

 The uncertain longevity of new technologies. According to BASF, the average process plant has a lifecycle of 30 years, compared with 10-15 years for automation hardware and just five years for software. A chemicals plant has to be certain that the technology will last without resulting in downtime.

 Confidence in cyber-security. As we are still in the nascent stages of IoT software, the chemical industry has been reluctant to fully adopt the technology. Despite potential productivity gains, in these early days, there is too high a cost of failure to decentralise decision-making, as security breaches raise the possibility of explosions, environmental damage and injuries or fatalities. Nevertheless, we expect PLM software to be an increasingly important technology for the industry. We also expect an increased adoption of IoT PaaS software; however, this technology is only likely to be used for data analytics/real-time surveillance rather than control actuators (which make decisions) using technologies such as wireless secondary sensors.

Wood & paper products – TAM up to US$50 bn We estimate that around 4% of global manufacturing fixed asset investment is into the metals industry, worth more than US$200 bn. The manufacturing process is fairly resource- intensive, involving products that have a low value density. This industry is already characterised by high software intensity – above that of automotive and electronics – and it is becoming increasingly automated, particularly in developed markets. We highlight IoT PaaS software and AGVs as the two most relevant technologies, where AGVs is the least penetrated of the two.

Textiles, leather & wearing apparel – TAM c.US$10 bn Accounting for only c.1% of global manufacturing fixed asset investment (worth c.US$50 bn), the textiles, leather & wearing apparel industry is the smallest industry we look at. Textiles, leather & wearing apparel is the most labour-intensive industry in our analysis, and so far, companies have typically responded to rising wages by moving to lower-cost locations. For example, we are seeing dramatic growth of textile industries in countries such as Cambodia, Bangladesh and Vietnam, as companies leave countries with rapidly rising wages, including China. However, given the very low level of automation, we expect the industry to benefit from labour-reducing technologies: cobots and AGVs. RFID and IoT PaaS are also important technologies.

Goldman Sachs Global Investment Research 44 April 13, 2016 Profiles in Innovation

Winners & losers

 Increasing factory floor digitisation leaves room for more players in the manufacturing arena, beyond the traditional capital goods companies  As new players emerge, we are likely to see more M&A, redefining the competitive landscape  New business models will also emerge, as localisation, customisation,

predictive management and control and monitoring are facilitated, but will

likely also drive deflationary pricing pressures in traditional hardware

Goldman Sachs Global Investment Research 45 April 13, 2016 Profiles in Innovation

New players, new business models

We expect more players to serve the market for future manufacturing technologies than the traditional capital goods equipment providers that have dominated the past decades. The boundaries between software/connectivity providers and hardware equipment are blurring. As this happens, we expect M&A to increase, as well as new business models to emerge through new aftermarket services and data analytics consulting. Regional balances in manufacturing are also likely to be challenged, with EMs that are pressured on cost leadership likely to adopt some of the newest technologies more quickly.

New players emerging The flip side of digitisation is that software allows for better optimisation of current asset bases by increasing utilisation rates, reducing the need for capital goods, and introducing a new set of formidable competitors: the SAPs and IBMs of this world. As the world of capital goods goes increasingly digital, the incumbents need to navigate a fine line between the opportunity and the threat of digitisation.

The opportunity sounds huge, but it is not free of challenges. Perhaps the most important one for capital goods companies will be to grow fast enough and reach sufficient size to offset the cannibalisation effect of better-optimised factories resulting in less demand for ‘hard’ equipment. For instance, increasing the utilisation of a fleet of compressors from 50% to 80% (through connecting it to an IoT PaaS) would reduce the number of compressors needed by close to 40%, but would not materially change the service needed and would involve substantial IT investments. Both servicing and IT investments are not necessarily only provided by the manufacturers of equipment.

Second, some of the FoF-enabling technologies, such as software, are not exclusively provided by traditional capital goods companies; indeed, the likes of IBM, Oracle and SAP are increasingly trying to penetrate markets traditionally served by electrical and machinery makers. Capital goods companies have the advantage of understanding how the physical assets of their customers work, but lag on the software side of engineering skills vs. the big IT providers.

Cisco expands IoT presence with cloud-based service platform provider Jasper

What is Jasper? In March, Cisco completed the acquisition of Jasper Technologies, a cloud-based IoT service platform, for US$1.4 bn. Jasper allows both enterprises and service providers to automate management of IoT services and drive monetisation on a global scale. Jasper currently has over 3,500 enterprise customers and works with 27 service providers in over 100 countries. The company targets a wide range of verticals including automotive, home automation, and manufacturing; customers include General Motors and Garmin. On the service provider side, per a 2014 Infonetics report, Jasper is among the key technologies supporting M2M platforms for AT&T, China Unicom, Telefonica, and Telenor. Jasper delivers its platform through a ‘Software as a Service’ revenue model.

What does it mean for Cisco? Cisco intends to wrap Jasper’s platform into a holistic IoT solution by layering in additional services including enterprise Wi-Fi, security for connected devices, and analytics to add-in device management. We view Jasper’s platform as complementary to Cisco’s existing IoT product suite and believe a combined offering will further extend Cisco’s IoT leadership. Recall that over the past several years Cisco has invested over US$1 bn in this emerging opportunity, which was generating US$2.4 bn in annual sales and growing at 45% per annum (per IoTWF in October 2014).

Goldman Sachs Global Investment Research 46 April 13, 2016 Profiles in Innovation

More corporate action to come Finally, it will be difficult to drive a meaningful earnings contribution from software and more connected assets in the near term purely through organic actions. This leaves M&A as one of the main levers companies are likely to use to tackle this opportunity. Several deals over the past few years in the capital goods sector highlight that there is already a focus on this: ABB acquired Ventyx and Mincom, two providers of enterprise software solutions for asset monitoring, in 2010/11; Schneider acquired Telvent, a provider of software for asset management, in 2011; Teradyne acquired Universal Robots in 2015; and Amazon acquired AGV maker Kiva in 2012. Within the sector, the ability to generate value from transformative deals has varied significantly by company. We believe the companies that can generate cash more quickly through their current portfolios will be better positioned to benefit from this inorganic opportunity; as tech companies enter the industrial space, we expect to see more M&A from industrials into tech as firms try to stay ahead.

Exhibit 46: Sector boundaries are blurring… Exhibit 47: …as US tech companies target manufacturing # of deals where cap goods companies acquire software # of deals where US tech acquirer targets industrial targets targets

60 5.0% 70 7.0% 4.5% 60 6.0% 50 4.0% 3.5% 50 5.0% 40 3.0% 40 4.0% 30 2.5% 30 3.0% 2.0% 20 1.5% 20 2.0% 1.0% 10 10 1.0% 0.5% 0 0.0% 0 0.0% 2009 2010 2011 2012 2013 2014 2015 2007 2008 2009 2010 2011 2012 2013 2014 2015 No. of US technology industrial deals (LHS) No. of software deals (LHS) Industrial deals as % of US Tech deals (RHS) Software deals as % of CapGoods deals (RHS)

Source: Bloomberg, Goldman Sachs Global Investment Research Source: Bloomberg, Goldman Sachs Global Investment Research

The emergence of new business models For equipment makers to succeed in the next phase of manufacturing, a first-mover advantage and leveraging the installed base will be key. Digitisation will affect companies in different ways. In an environment with less growth and more EM competition, installed bases will increasingly be the key competitive advantage for global capital goods companies. Those that already have a high penetration of services and maintenance in their sold equipment have a quicker lead into understanding what revenue streams can be levered/protected by better device connectivity. Taking history as a guide, we believe that whether digitisation turns out to be an opportunity or a threat will ultimately reflect each company’s underlying ability to reinvent itself. We see companies that have a good knowledge of their installed base and a superior ability to redeploy capital as better positioned (as highlighted in the Competitive Positioning metrics presented in our report Preparing for the next industrial revolution, April 28, 2014). Which players are able to capture the benefits of the exponential creation of data in the industrial space remains to be seen, but three benefits appear almost certain for those that can:

Goldman Sachs Global Investment Research 47 April 13, 2016 Profiles in Innovation

 Controlling data will probably mean more customised value-add client interactions. Data gives companies capacity to be more proactive in addressing customer issues.

 More frequent client interactions most likely mean more resilient revenue streams…

 ...while more customised value-add client interactions usually mean better margins and returns.

Exhibit 48: Digitisation amplifies recurring revenues Exhibit 49: Companies that succeed in leveraging Recurring revenues as % total for our European software vs. digitisation to create recurring revenues will benefit from capital goods coverage (2013) greater stability… Service/aftermarket estimated peak-to-trough sales growth over the past 15 years

70% 90pp

59% 80pp 60% 51% 70pp 50% 60pp

40% 50pp

40pp 30% 25% 30pp 20% 20% 20pp Peak-to-troughSales Growth Gap % Sales from captive business 10% 10pp

0pp 0% Alstom Atlas FLSmidth Kone Metso Schneider Wartsila Weir - Average Median (GE) - Copco Electric Minerals Thermal Software Traditional Capital Goods Power Aftermarket/Service Products

Source: Company data, Goldman Sachs Global Investment Research Source: Company data, Goldman Sachs Global Investment Research.

As machines take over, business models will need to evolve quickly. In developed markets, the penetration of PCs, tablets, and smartphones is now almost complete, with 75% of people in the United States having access to mobile broadband internet. This phenomenon has changed the way we work, communicate and consume so much that it is difficult to remember what it was like in the 1990s. As unintuitive as it sounds in the era of Facebook, Google and SAP, the vast majority of capital goods companies still interact with their installed base and customers in the same way we interacted with each other in the 1980s. For the majority of industrial companies, if equipment needs servicing, customers will call an onsite technician. But this is all about to change. What the IT revolution has done to how humans work and communicate over the past 20 years is now happening to machines; and it has the potential to be as revolutionary to how machines operate from day to day as it has been for us. In our view, this is likely to disrupt many business models for the companies that produce and maintain these machines, creating business opportunities for the companies that best utilise the data and, just as it has for consumers, imply deflationary pressure on “like-for-like” products. Innovation will need to be faster and business models more agile, and, most importantly, the way capital goods companies interact with their customers will have to change.

However, when we look back historically, multi-industry capital goods companies have a track record of reinventing themselves and some of them are already on that path again with their latest digitisation-related deals (e.g. Schneider, Siemens and GE).

Goldman Sachs Global Investment Research 48 April 13, 2016 Profiles in Innovation

IoT allows Kaeser Kompressoren to shift to a more competitive equipment-as-a-service business model

The challenge: Like any traditional manufacturer of capital goods equipment, Kaeser was searching for incremental revenue drivers to optimise its capacity in the face of slowing end markets.

The solution: A few years ago, it began incorporating sensors in its compressors and examining the data it collected in a SAP IoT platform as a potential source of incremental revenue. It started by building a predictive maintenance programme, but ultimately it completely reinvented its way of selling its products while simultaneously increasing its customer base: selling compressed air, instead of air compressors. This shift opened up the compressor market opportunity to smaller customers for which the decision to buy compressors would mean too much of a capital commitment; instead, these customers can tie their investment to their revenues (i.e. pay for the air as they use it). Sensors and big data made it possible for Kaeser to build the platform for this new business. Furthermore, this new business model encouraged the company to be more cost conscious: as the machines are no longer the revenue- producing element and customers pay for uptime, having faultless compressors that do not require any onsite service is the primary goal (i.e. Kaeser makes material savings on service staff).

The expected winners – large installed bases; data analytics leaders We do not have a clear-cut answer on whether hardware or software providers will win the race to capture value from higher automation. What is clear to us is that hardware makers with large installed bases will have a privileged position and software providers with big data analytical expertise can make meaningful inroads. In the short to medium term, the need for equipment to upgrade capex may benefit the hardware makers, but we believe that over the long run, more of the technologies profiled will make the manufacturing system more optimised, longer lived and ultimately in need of lower capex.

Leading hardware makers could leverage their large installed base and provide production process/product optimisation services based on big data collected from their equipment. Most major Western capital goods companies are already ramping up their software teams and R&D spending in order to be prepared for this.

The largest beneficiaries will likely be the key component makers for enabling input technologies, such as those in the semiconductor/sensor industry, as the billions of machines to be inter-connected will create a substantial opportunity for their product. Providers of integrated Enterprise Resource Planning and Manufacturing Execution System software should also benefit from more data being available, which will drive upgrades of capacity of existing systems.

Where will it be built? Challenging regional establishments The past three decades have seen unprecedented levels of globalisation as companies from developed economies relocated their manufacturing practices to lower-cost emerging markets.

We expect the future of manufacturing to be less labour-intensive, although more skilled labour will be needed. Speed to market will become increasingly important and we expect transport costs as a percentage of total costs to increase. Therefore, some of the motives that fuelled globalisation in recent years are no longer likely to influence decisions. Instead, we expect a critical factor to be manufacturers’ need to be closer to their customers. Does this mean we will see a wave of reshoring? Not necessarily, as most of the growth in new customers is still coming from emerging economies.

Goldman Sachs Global Investment Research 49 April 13, 2016 Profiles in Innovation

China: The factory of the world China is still the largest major manufacturing centre in the world, representing 30% of global manufacturing GDP. Chinese companies have been talking extensively about Industrie 4.0, but we think most companies have yet to complete the 3.0/automation stage. Shanghai Highly Group, for example, one of China’s largest compressor manufacturers, started to employ Industrie 4.0 in its production process, but a lack of big data analytical capacity/platform has limited its efforts to move towards intelligent administration and intelligent production.

Automation is still low in China, and while the labour cost gap is narrowing, the country still benefits from significantly less expensive talent than in DMs, suggesting a continued incentive for a slow conversion of manufacturing from manual to automated. In our analysis of customer budgets, we note that the capex-to-sales ratio of manufacturing companies is still much lower than that of the mature DM manufacturing companies.

Exhibit 50: Robot penetration varies markedly even in autos Exhibit 51: China’s technology gap may be narrowing in Robots per 10,000 workers vs. cars produced in 2014 electrical engineering but mechanical engineering is still far behind Patent Cooperation Treaty application by country

Mechanical engineering Chemicals Electrical engineering Others 70,000 1600 25mn

1400 60,000 China Germany Japan US 20mn 1200 50,000 1000 15mn 40,000 800

10mn 600 30,000

400 5mn 20,000 200

0 0mn 10,000 Japan Germany USA France Italy Indionesia China Turkey India

Number of industrial robots per 10,000 persons employed in the automotive industry 0 (LHS) Cars produced (RHS) 2004 2006 2008 2010 2012 2014 2004 2006 2008 2010 2012 2014 2004 2006 2008 2010 2012 2014 2004 2006 2008 2010 2012 2014

Source: IFR, IHS Global Insights, Goldman Sachs Global Investment Research. Source: World Intellectual Property Organization.

What about the west? The US and Germany continue to lead several new technological developments, although even in these areas China is catching up. As our GS SUSTAIN team argue in their report Germany AG: Don’t look back (September 16, 2015), corporate Germany’s leadership in ‘premium hardware’ has been a key driver of its success over the last decade. However, the “premium” element of hardware will increasingly be defined by the software and servicing elements attached to it. This transition introduces new and formidable competitors, changes what constitutes a successful business model, and more broadly puts Germany’s competitive advantage “up for renewal”. In its core areas of strength, autos, capital goods and consumer durables, Germany’s corporates need to lead this technological transition in order to sustain their leadership. The rise and fall of Germany’s consumer electronics industry provides an example of the risks involved in not keeping pace with disruptive change.

Goldman Sachs Global Investment Research 50 April 13, 2016 Profiles in Innovation

Asia’s Automation Future

Over March 9-29, we hosted a factory automation tour, visiting 26 factory automation-related companies in Japan and China. Below, we highlight examples of how manufacturers in Asia, previously renowned for low labour costs, are in the midst of rapid change to become the factories of the future.

Best in class of Asia part 1: New Chinese factories are among the most automated in the world

China is now experiencing the kind of rapid change that previously forced Japan and other developed economies to increase factory automation. The pace at which it is happening in China is eye-catching. In addition to rising personnel costs and other fixed costs, automation is being driven by demand for higher quality and precision, and limits to economies of scale.

China’s auto industry is the most conspicuous example. Typically speaking, the auto industry has the highest automation rate (robot penetration) of all industries. What is noteworthy is that the new auto plants in China actually have the highest penetration rate of robots across the globe.

For example, Nissan’s Dalian plant, its newest plant in China (production started in October 2014), has a robot penetration rate of 83%, the highest of any Nissan plant in the world. By contrast, the robot penetration rate at Nissan’s Huadu plant, which started production in May 2004, is still around 56%. Local company Geely Auto is aggressively automating/installing the newest machines in stamping, welding, and painting processes at its new Ningbo plant, which was undergoing renewal to make the new model “Pore”. The reason for the aggressive automation on these processes is that these are critical determinants of vehicle performance. Geely Auto has installed the newest foreign capital goods/machines instead of Chinese local machines. The plant has monthly production capacity of around 150,000 units and uses more than 100 spot welding robots; a ratio on a par with that of foreign automakers’ plants.

Best in class of Asia part 2: Penetration is uneven and initial investment is still large

Automation penetration in China’s electronics industry is very uneven. Shortening return on investment periods and decreasing fixed-cost burdens are key factors in the semiconductor fabrication space. Having the newest SPEs/fully automated production lines is critical to make the latest/most advanced nodes, and manufacturing in China is no exception. Yet, the rate of automation is much lower on high-volume products with wider processing nodes (for devices used in autos or conventional electronic applications) and among Chinese companies, as they try to balance the fixed cost burden by leveraging manual labour/assemblies and second-hand SPEs.

Chinese electronic manufacturing service (EMS) companies have strengthened their market presence in recent years. The level of automation in EMS plants is on an altogether different level to that of traditional industries, particularly for smartphones, where annual production runs to several tens of millions of units. Hon Hai, China’s largest EMS company, procures thousands or even tens of thousands of production machines from different FA companies every year, i.e. the number of robots in a single-product new auto plant is typically in the hundreds. In contrast, an EMS may have to introduce thousands more robots to make a new product.

Best in class of Asia part 3: Factories of the future may appear sooner than expected

Given China’s supportive policies, we think a factory of future may appear in China sooner than expected (even though it may not be justified economically yet). A leading system integrator has stated that it is in discussion with a provincial level government industrial investment fund (Rmb20 bn AUM) to build a highly automated pilot factory, fully funded by the aforementioned fund, to set an example for the manufacturing industry for that region. The factory is supposed to include hardware and manufacturing software that is as advanced as possible and may become a general manufacturing services provider, or a factory that manufacturers in the region can rent in a flexible manner.

Goldman Sachs Global Investment Research 51 April 13, 2016 Profiles in Innovation

28 companies with exposure to the Factory of the Future

In this section we highlight a non-exhaustive group of companies exposed to the highlighted technologies as well as some of their key suppliers.

Covered by Samuel 3D Systems (DDD, Neutral): Metal platform a key differentiator Eisner, US SMID Capital Goods analyst Company profile DDD is a manufacturer and designer of additive manufacturing printers, materials, and services, with a broad platform and historically acquisitive strategy. DDD focuses on offering a full line of diverse printer types to service the additive manufacturing market. Over a roughly 30-year history, DDD has pioneered many 3D printing technologies, and through an aggressive acquisition strategy, has focused on developing an end-to-digital manufacturing solution it calls the “digital thread”, allowing it to offer an extensive printer, maintenance, software, healthcare, and printing service platform.

Exposure to the Factory of the Future DDD is a global leading manufacturer of 3D printing; its technologies allow for small lot sizes, more flexible manufacturing, and more complex manufacturing at lower costs. In addition, we believe DDD has an advantage with respect to its ProX metal printing platform, which it acquired in 2013 and has developed since then. We believe the industrial market is likely to be a bigger end market for participants relative to the consumer/education markets, and that a metal printing platform should offer a long-term competitive advantage. That said, material knowledge needs to increase in order to spur greater adoption. DDD also offers a more extensive suite of accessory products relative to competitors, including design software, haptic control devices, handheld scanners, and other products.

Covered by Daniela ABB (ABBN.S, Sell): Exposed to the shift to human-robot collaboration Costa, European Capital Goods analyst Company profile ABB is a multinational group operating in the power and automation technologies sector. It operates in around 100 countries, employing about 145,000 people. Its Power division focuses on transmission & distribution equipment and power plant automation, serving electric, gas and water utilities, as well as commercial and industrial customers. The Automation business serves a full range of industries with measurement, control, protection and process optimisation applications, including industrial robots, where it has a global market share of 10%-15%.

Exposure to the Factory of the Future ABB is a key enabler through its robotics business, which makes up c.5% of group revenues. In particular, the group launched its dual-arm collaborative robot YuMi in 2015. This product is extremely accurate, able to learn, and is connected to ABB’s global remote condition monitoring centre. It can be integrated into production lines without the need to redesign the space. However, ABB’s offering extends beyond just robots to software, complete systems, fully integrated application equipment, manufacturing cells and a comprehensive offering with IoT integration. ABB also offers additional services such as remote monitoring and failure prediction, and remote services robot back-up. ABB’s most recent acquisition in this space of collaborative robotics was Gomtec, a company developing mechatronic systems combining mechanical, electrical, telecommunications, control and computer engineering for a diverse range of customers.

Goldman Sachs Global Investment Research 52 April 13, 2016 Profiles in Innovation

Covered by Willy Chen, Advantech (2395.TW, Neutral): Key supplier of embedded Taiwan Capital Goods components for industrial IoT systems analyst Company profile Advantech is the global leader in embedded computer technology that is used to gather data and process information in various end markets. Advantech is one of the few vendors globally that has the capability to design from computer board & modules level to application ready. This allows its downstream system integrator clients to easily embed their products.

Exposure to the Factory of the Future Advantech provides the gateways, networks, servers, visualisation and data management products to connect and control the nodes (machines) in the Factory of the Future. Advantech’s embedded products can be installed in the machines and make the manufacturing process more intelligent and connected. In order to make its products compatible with most devices globally, Advantech also partners with Intel and Microsoft and provided the first IOT Gateway Development Integration Kit in March 2016 to accelerate the adoption of smart factory (Industrial IoT). In addition to providing the hardware products, Advantech also develops the IoT software “WebAccess”, which uses the Microsoft Azure cloud platform to process comprehensive data gathered from Advantech devices in the factory. Through the software and the cloud platform, users can benefit from highly efficient management and fast interconnectivity among these cloud services.

Covered by Simona Cisco (CSCO, Buy): Connecting the Internet of Things Jankowski, CFA, US Hardware & Company profile Communication Cisco is a leading global provider of networking technology, including systems, software and Technology analyst services, with US$49.6 bn in revenues in CY15. The company has been a long-standing leader in the US$24 bn switching industry, with about 62% market share, and switching products drove 30% of its total revenues in CY15. Cisco’s second-largest segment, routing, drove 15% of total revenues, with the company having a share of about 40% in the US$10 bn service provider routing market and 75% in the near-US$3.2 bn enterprise routing market. In addition to its core switching and routing revenues, Cisco has a leading position in a number of other categories, such as service provider video, which includes video infrastructure and software (US$3.4 bn in CY15 sales), collaboration (US$4.2 bn in CY15 sales), Wi-Fi (US$2.6 bn), servers (US$3.4 bn), and security (US$1.8 bn). Cisco’s services business, which generated sales of US$11.5 bn last year, accounted for the remaining 23% of CY15 revenues.

Exposure to the Factory of the Future Cisco is aggressively pushing into the IoT, with over US$1 bn of organic investment supplemented with the recent US$1.4 bn purchase of Jasper Technologies. Jasper is a leading cloud-based IoT service platform for both enterprises and service providers. The company currently has more than 3,500 enterprise customers and works with 27 service providers in over 100 countries. More broadly, Cisco’s approach to the IoT is focusing on the industrial vertical rather than consumer markets, as that is where it sees the most value and therefore the largest potential profit opportunity. Targeted industries include manufacturing, transportation, smart/connected cities, oil and gas, and smart grid.

Goldman Sachs Global Investment Research 53 April 13, 2016 Profiles in Innovation

Covered by Dassault Systemes (DAST.PA, Buy): PLM software market leader at Mohammed Moawalla, the heart of new industry expansion European Software analyst Company profile Dassault is a global software vendor that operates in the PLM (Product Lifecycle Management) space, offering an integrated suite of software solutions across industries: this accommodates the entire product lifecycle, from 3D design to mock-up, rendering, collaboration, simulation and digital manufacturing. While historically Dassault’s core end- markets have been automotive, aerospace and industrial equipment, over the past few years, the company has expanded into industries such as consumer packaged goods, retail, high-tech, construction and life sciences, which now represent c.30% of the group’s software revenues (2015). Dassault has a total customer base of 190,000 and sells its solutions in more than 140 countries. As of December 2015, Dassault had c.14,000 employees.

Exposure to the Factory of the Future In our view, as the market leader in PLM software, Dassault remains a key beneficiary of industry megatrends, such as the Internet of Things (IoT) and Factory of the Future, which should drive pent-up demand for reinvestment to transform supply chains. Dassault has developed, both organically and inorganically, a broad portfolio of PLM solutions encompassing offerings such as CAD, collaboration, simulation and digital manufacturing, catering to both OEMs and the supply chain. Its 3DExperience platform further enhances its ability to provide a more holistic PLM offering. While historically, Dassault has had a focus on automotive, aerospace and industrial equipment (c.60% of 2015 revenue), it has significantly increased its addressable market by moving into adjacent product verticals and diversified industries. Furthermore, 3DExperience will allow Dassault to create packaged industry applications for specific processes, along with providing a common infrastructure and thus tailoring it to specific vertical industry requirements.

Covered by Willy Chen, Delta Electronics (2308.TW, Buy): Power behind the factory Taiwan Capital Goods analyst Company profile Delta Electronics was founded in 1971 and has been the global leader in switching power supply since 2002 and DC (direct current) brushless fans since 2006. Delta has 38 production plants, 60 R&D centres and 153 offices across the world to provide rapid service to its clients. Delta’s product portfolio includes power supply for consumer electronics, system power for telecom and renewable energy, electronic components (thermal & magnetic products) and industrial automation products (motion control & robotics). Delta has been included in the Dow Jones Sustainability™ World Index (DJSI World) since 2011 and was the only company from Greater China to be named in the Climate Performance Leadership Index (CPLI) of the 2014 Carbon Disclosure Project (CDP).

Exposure to the Factory of the Future Delta launched its SCARA and articulated robots in 2015 and is planning to launch lightweight articulated robots and parallel robots in 2016/17. With its capability to produce the key components internally (controller, servo motor), Delta is able to provide competitive robot products to its clients. Delta is the third-largest producer in China’s servo motors/drives market (11% share). Aside from the robotics business, Delta also provides power systems to datacentres and telecom base stations, which supply stable power to support increasing demand for data processing in the Factory of the Future. Delta now holds the global No.1 share (17%) in telecom power systems. For datacentres, Delta has developed a modularised solution “Delta InfraSuite”, which provides a one-stop-shop service, including power management (UPS and PDU), accessories, precision cooling and environmental management.

Goldman Sachs Global Investment Research 54 April 13, 2016 Profiles in Innovation

EOS GmbH (Private, Not Covered): Leader in metal printing

Company profile EOS is a leader in sintering technology, and is headed by Dr. Hans Langer, its founder and CEO. Following its establishment more than 20 years ago, the company expanded from the production of SLA and SLS systems to the production of metal sintering platforms in the late 1990s. Since the late 1990s, the company has focused exclusively on the production of powder-based sintering. We believe EOS’s key competitive differentiation is achieved through its expertise in laser optics and their influence on metal melt and print repeatability.

Exposure to the Factory of the Future Additive manufacturing (AM) providers such as EOS will play a role in the Factory of the Future by allowing for greater part complexity, smaller lot sizes, and faster prototyping and product development. In particular, EOS focuses on a sintering-based platform, and leads the aerospace and medical fields with its large format metal printers. One of EOS’s largest customers is GE, which uses the company’s products in manufacturing its new LEAP engine. In addition to the more customisable products that additive manufacturing allows, we also believe that smaller lot sizes enabled by AM should allow for faster cycle times and smaller capital investment for certain components of the supply chain. For instance, the high-grade metal printing platforms offered by EOS could allow for leaner supply chains and less inventory of replacement parts, as small footprint printers coupled with just-in- time delivery of metal powder inventory could replace their larger investment peers.

Covered by Yuichiro Fanuc (6954.T, Buy): Global leader in industrial robots and CNCs Isayama, Japan Machinery analyst Company profile Fanuc was founded in 1972 and is the global leader in the industrial robot and CNC (computerised numerical control, the operating system of industrial machines such as machine tools/plastic injection molding machines etc.) markets. Fanuc holds a global market share of 17% in industrial robots, and more than 50% in CNC. Historically, its operating margins (35%-40%) and CROCI (30%) have been among the highest in the global industrials sector. The company has been able to maintain exceptionally high margins/returns thanks to: (1) full integration of its key components (controllers, amps, motors); and (2) a standardised and fully automated production system, which is often represented by the term “robots making robots”.

Exposure to the Factory of the Future Despite its dominant industry position, Fanuc has long been unable to launch new products in entirely new sectors. However, its mass marketing of collaborative robots, establishment of a fibre laser joint venture, and other initiatives to tap into new business sectors suggest the company is changing. We think fully fledged development of collaborative robots, in particular, reflects Fanuc’s efforts to generate new added value. Collaborative robots, which are designed to work in the same environment as humans, represent the opposite end of the spectrum to Fanuc’s founder’s concept of creating separate working environments for humans and robots. Fanuc’s collaborative robots are also very different in appearance, covered in a green sponge jacket as opposed to Fanuc’s traditional yellow colour. We see this as a clear example of a change in Fanuc’s stance, aimed at promoting products to a new customer segment.

Goldman Sachs Global Investment Research 55 April 13, 2016 Profiles in Innovation

Covered by Joe Ritchie, General Electric (GE, Neutral): Driving force behind the Industrial US Multi-Industry analyst Internet concept Company profile GE is the biggest industrial player globally, with US$100 bn+ in annual revenues. GE has dominant positions in key end markets such as power generation, jet engines, locomotives, healthcare and O&G. In addition, GE has an extremely strong services model, as evidenced by US$200 bn+ in services backlog that provides stability to an otherwise cyclical portfolio.

Exposure to the Factory of the Future GE has been at the forefront of IoT, investing in and developing emerging technologies such as additive manufacturing, data analytics and predictive maintenance. Over the long term, we see GE as one of the best positioned companies to take advantage of these technologies given its head start, a vast installed base and continued spending to develop capabilities (currently c.US$1 bn run rate). Specifically, key GE initiatives that underpin our view are: (1) Controls & sensors: several of GE’s products such as HDGTs, jet engines and locos are big-ticket items that have long usable lives, high operating/maintenance costs and generate a lot of data. GE has been one of the pioneers in developing sensors that can work in hazardous environments (e.g. ceramic matrix composites) and are able to provide detailed/real-time information; (2) Predix is the cloud platform for IoT (recently made open source), which developers can use as the foundation for building apps that help businesses optimise assets, fleets, and overall operations. Predix combines distributed computing, big data analytics, asset data management, machine-to-machine communications and mobility. The platform extends beyond GE assets to connect machines from any vendor or vintage.

Covered by Yuichiro Harmonic Drive Systems (6324.T, Buy): Global top share company Isayama, Machinery in small precision reducers for cobots analyst Company profile Harmonic Drive Systems was established in 1970 and since its inception has concentrated on the production of high-precision speed reducers (73% of sales). In 2014, HDS appointed Akira Nagai, formerly of Mitsui & Co., as its new president, unveiling a long-term vision targeting sales of ¥50 bn by FY2020.

Exposure to the Factory of the Future Speed reducers are an important component of collaborative robots. With a global market share of roughly 90% in speed reducers used in small assembly/transfer robots with maximum load capacity of less than 6 kg, HDS is primed to benefit from the growth of cobots in manufacturing. However, outside of manufacturing, we also see the service robot market, with applications in nursing care and other services, as a promising source of demand over the much longer term.

Covered by Jacqueline Han’s Laser (002008.SZ, CL-Buy): China’s leading laser solution Du, China Machinery provider; a wide set of potential new applications analyst Company profile Han’s Laser was established in 1996 and is now the No.1 laser equipment manufacturer in China, with a 37% market share in smaller power equipment thanks to its strong market access and cost advantage. The company has more than 200 machine models (including the laser engraving/marking series, welding series, cutting series, sub-surface engraving series and display series) and its equipment has been widely used in the manufacturing of high-precision and high-consistency products such as smartphones, lithium-ion batteries and semiconductors. Some of Han’s key customers include smartphone makers Apple, Samsung, Huawei, Xiaomi and local lithium-ion battery makers. Han’s Laser could potentially produce full in-house fibre laser, which may potentially further expand the company’s margins.

Goldman Sachs Global Investment Research 56 April 13, 2016 Profiles in Innovation

Exposure to the Factory of the Future Laser application is the icing on the cake for automation upgrades, as it is more stable, flexible and easier to integrate into an automated production line. Han’s Laser has comprehensive exposure to FoF in the following regards: (1) Integration of factory automation with laser technology, such as 3D laser welding robot stations. With a laser processing head and pre-set laser beam, the laser-robot can provide accurate multi- dimensional welding depths/angles and capabilities. Currently, welding robots are the second-largest industrial robot type, accounting for 25% of total robot sales in Japan as of 2014. (2) Han’s Laser established its metal 3D printing centre in December 2015. 3D printing/additive manufacturing is a digital process using fibre lasers for metal melting or sintering. Currently, the process is constrained by cost and material, but we expect the global market to expand significantly over the next 15-20 years. (3) Potential new applications include machine vision and LIDAR for autonomous driving and automated guided vehicles.

Covered by Joe Ritchie, Honeywell (HON, CL Buy): Automation player leveraging a large US Multi-Industry installed base analyst Company profile HON is an industrial conglomerate with c.US$39 bn in annual revenues and diversified end markets (resi, non res, O&G, aero and auto). HON maintains a strong position across key end markets/product lines, which, coupled with the company’s internal repositioning initiatives, has resulted in best-in-class execution over the past few years.

Exposure to the Factory of the Future Since 2014, HON has greatly increased its focus on several fast-growing trends such as connected homes and data analytics. HON has navigated through the evolving industrial landscape by: (1) Introducing initiatives like HUE to drive faster product development; (2) increasing its emphasis on software capabilities, with a higher proportion of headcount additions and M&A spend going towards this; and (3) leveraging its vast and diversified installed base (130,000 aircraft, 100 mn vehicles, 150 mn homes etc.) to drive increased data collection; this, coupled with HON’s analytics and cloud services, should help with predictive maintenance, improved productivity and higher customisation. In addition, HON has a strong position in RFID (the number 2 player globally in scanning & mobility) and has bolstered its portfolio through the acquisition of Intermec. HON’s products include RFID readers, printers, tags, smart labels and mobile computing solutions. Specifically, HON’s range of Optimus mobile computers helps customers realise the benefits of RFID in a dynamic manner (including data collection), enabling better supply chain and inventory management, and allowing for real-time decision making.

Jacqueline Du, China Inovance Technology (300124.SZ, Buy): China's top-class motion Machinery analyst control supplier

Company profile Shenzhen Inovance Technology is one of the most innovative industrial motion control suppliers in China. Inovance launched its first product, low-voltage inverters, back in 2004, and became the No.1 domestic player in China’s servo market in 2008. Currently, the company’s products are applied in a wide range of downstream industries, including elevator control systems for elevator OEMs, and EV controllers for Yutong Bus (one of the largest EV bus makers in China). The company also supplies a broad range of factory automation products (including servo, PLC and sensors). Inovance has grown its business organically through expanding its strong R&D workforce, while actively investing in emerging areas of technology. The company invested in its machine vision subsidiary in 2013, and has become the domestic pioneer in this field.

Goldman Sachs Global Investment Research 57 April 13, 2016 Profiles in Innovation

Exposure to the Factory of the Future We believe Inovance’s products are indispensable cornerstones for the China chapter of the FoF revolution. The company has a footprint in the competitive motion control markets, mostly dominated by European and Japanese automation giants such as Siemens/Schneider and Mitsubishi/Yaskawa, while enjoying a strong domestic advantage underpinned by market share gains from providing customised products to niche markets, and being relatively more nimble than foreign peers. Inovance has maintained a solid track record of launching a new product every 1-2 years and quickly becoming the domestic leader in that field.

Covered by Yuichiro Keyence (6861.T, Neutral): One-stop FA sales-shop/consultant, with Isayama, Japan exceptionally high margin/returns Machinery analyst Company profile Founded in 1972, Keyence is a leading maker of automatic control equipment, electronic equipment and FA sensors. Keyence handles design, development and sales in-house and outsources all production to Japanese companies. Keyence is well known for a consultative sales approach in which it works closely with customers to determine how its products can improve their productivity. In this sense, its business model is similar to that of a consulting firm—it is selling its ability to add value. Keyence has one of the strongest profitability measures of any listed machinery company in the world, including a gross margin of 80%, an operating margin of more than 50%, and a high CROCI.

Exposure to the Factory of the Future As a consulting-oriented company, Keyence’s competitiveness depends on its personnel. Development and service expertise have enabled Keyence to consistently deliver highly innovative products over a long period. While Keyence supplies products ranging from 3D printers to PLCs, the sensor business – an area of strength since foundation – is the crown jewel. Keyence has a broad sensor lineup and along with Cognex (US), it is well positioned to benefit from the growth in machine vision driven by increasingly complex inspection processes.

Kion (KGX.DE, Not Covered): Staying ahead of the game in AGVs

Company profile Kion is the world’s second-largest provider of forklift trucks and other material handling equipment (including autonomous equipment). The group has sales of over €5 bn and Europe is its largest market.

Exposure to the Factory of the Future In 2015, Kion acquired Egemin, an automated logistics provider with operations that include developing and manufacturing automated forklifts and automated guided vehicles; these significantly reduce labour costs in factories and logistics centres (labour accounts for 80% of the total cost of owning a forklift). In addition, its traditional trucks are becoming more connected through the use of software, enabling higher utilisation rates within factories, the monitoring of driving quality, and to an extent predictive maintenance.

Kuka (KU2G.DE, Not Covered): Pure-play automation company at the forefront of collaborative robotics and AGVs

Company profile Kuka is a manufacturer of industrial robots and provides solutions for factory automation. The company is organised into three main operating segments: Robotics (30% of sales; the group has a c.14% global share in this market), Systems (49% of sales; this involves integrating automation equipment and helping assemble production lines) and Swisslog (21%; automated logistics equipment). Group sales come almost entirely from manufacturing industries, with the exception of a small proportion that comes from healthcare and warehouse distribution logistics (mainly through Swisslog).

Goldman Sachs Global Investment Research 58 April 13, 2016 Profiles in Innovation

Exposure to the Factory of the Future Kuka, through its iiwa (intelligent industrial work assistant) range, is at the forefront of collaborative robotics. The group estimates that it has a market share of around 10%-15% and expects not just increased cobot penetration, but increased use of mobile cobots. Mobile cobots are enabled through the use of the AGVs that Kuka develops internally, with this business strengthened in 2015 through its acquisition of Swisslog. Kuka is a pure-play factory of the future enabler and is currently the leading robotics provider (by share) in the automotive industry; however, the group is pushing forward with expansion into other industries (in 2015, general industry robotics orders grew 19%, surpassing automotive, which grew 4% yoy).

MJ Intelligent System (Private, Not Covered): Innovating China’s “smart factory”

Company profile MJ Intelligent System (MJ) is a Chinese factory automation system integrator, servicing a wide group of industries, now including auto/auto parts, home appliances, machinery and metal processing. MJ helps form “smart factories” by leveraging its proprietary SCADA (supervision control and data acquisition), MES (manufacturing execution system), WMS (warehouse management system) and big data analytics software. Some of its major clients are global/domestic leaders in their respective industries. The company has been selected to participate in the drafting of five industry standards under China’s “Manufacturing 2025” policy initiative.

Exposure to the Factory of the Future MJ’s “smart factory” goes beyond simply helping clients add robotic arms to the production line. Compatible with most major brands’ CNC and PLC, MJ collects and integrates real-time operating information from previously segregated automation equipment on the factory floor and communicates it with MJ’s self-designed Manufacturing Execution System. This helps the factory achieve flexible manufacturing through its powerful production process simulation and to connect equipment such as AGVs. Taking advantage of the information collected on the machine, MJ’s big data software monitors the status and efficiency of each working station and helps the factory maximise utilisation through predictive maintenance, as well as automatic adjustments. Thanks to MJ’s SCADA compatibility, a “smart factory” can add different brands equipment/sensors to the production process quickly and operating data can be fed back for further optimisation

Covered by Daiki Nidec (6594.T, CL Buy): Positioned to benefit from motor and drive Takayama, Japan system demand for robots/automation Electronic Components analyst Company profile Nidec was established in 1973 in Kyoto by current president Shigenobu Nagamori and three associates. The production of brushless DC motors started in 1975. In the 1990s Nidec established a leading global share in HDD spindle motors on the back of expanding PC demand. The company is renowned for its strategy of growth through M&A and has accelerated its overseas M&A related to medium-sized and large motors for auto and home appliance/commercial/industrial (ACIM). Products for auto and home appliance/commercial/industrial applications already account for around 46% of total sales (roughly half each), and we expect these areas to drive growth over the next decade, backed by technological trends.

Goldman Sachs Global Investment Research 59 April 13, 2016 Profiles in Innovation

Exposure to the Factory of the Future In the home appliance/commercial/industrial domains (23% of total sales), Nidec is focusing on new applications broadly related to the Internet of Things (IoT)—such as smart factories, robotisation and automation of warehouses, automated guided vehicles (AGV), cleaning robots, and high-performance drones. For instance, subsidiaries NMC (formerly the motors division of Emerson), Nidec Shimpo (transmissions and speed reducers for drive systems), and Nidec Kinetek (motor control technology) collaborated together to win orders for a warehouse management system from a major US online retailer. We expect this business to expand to ¥50-60 bn over the next few years (we think Nidec began deliveries in earnest from the start of 2016). We believe the company is currently in discussions and fielding many enquiries related to motors and systems for AGVs, cleaning robots, and high-performance drones.

Covered by Joe Ritchie, Rockwell Automation (ROK, Neutral): At the forefront of discrete US Multi-Industry analyst automation Company profile ROK is one of the global leaders in automation, particularly on the discrete side. Key products include controllers, plant management systems, electronic operator interface devices, visualisation software and intelligent motor control products. Over time, the company has consistently demonstrated the ability to adapt to emerging trends/technologies, which has resulted in top-line outperformance vs. peers over the cycle.

Exposure to the Factory of the Future ROK was one of the first large industrial companies to talk about the “connected enterprise” and has been in a partnership with Cisco for almost a decade, co-developing products for IoT applications. Given Cisco’s leadership in Ethernet and security, and Rockwell Automation’s familiarity with interface on the factory floor, the partnership seems advantageous for both. ROK has also been one of the earliest to recognise the convergence of capex and opex, as well as IT and machines, and has been working towards more integrated solutions. To this end, ROK has developed a powerful web-based manufacturing platform as a service that integrates all data into a single information management and decision support system. The system is designed for industrial environments, and can connect with existing PLCs (like ROK’s Logix) and databases.

Covered by SAP (SAPG.DE, Buy): Cloud products and S4HANA to enable factory Mohammed Moawalla, European Software of the future analyst Company profile SAP is a global enterprise software company which provides its customers with real-time, cloud-based products and solutions. The company is a leader in business applications and analytics software, especially Enterprise Resource Planning (ERP) and Supply Chain Management (SCM). SAP's product strategy centres on applications (on-premises, cloud and hybrid), platforms (database, analytics and platform as a service (PaaS)) and business networks. SAP is investing heavily in developing its cloud product portfolio and in-memory product offering: HANA. Its recently launched application offering, S/4HANA, is core to SAP’s innovation. SAP has exposure to a diversified range of verticals with consumer, manufacturing and energy/natural resources accounting for c.65% of 2015 revenues. SAP has over 300,000 customers and S4 HANA is already gaining traction with 3,200+ customers (end 1Q 2016), the group operates in >180 countries with c.77,000 employees.

Goldman Sachs Global Investment Research 60 April 13, 2016 Profiles in Innovation

Exposure to the Factory of the Future SAP has a strong entrenched position with manufacturing companies, especially in ERP, SCM and Supplier Relationship Management (SRM). We see SAP as a key beneficiary of emerging mega trends like IoT(Internet of Things), which require connecting and collaborating mechanical, electrical and software elements in different industry segments. Further, the shift to cloud as a means for applications delivery is driving investments in integration, middleware, monitoring, security, and analytics technologies to enable the factory of the future. We believe SAP’s strong applications footprint around sourcing, supply chain, logistics and procurement coupled with strong analytics capability, will allow it to develop next generation applications that enable the factory of the future. These will be built on its S4HANA platform that facilitates real-time processes and can be deployed either on premise, in the cloud or in a hybrid environment.

Covered by Daniela Schneider Electric (SCHN.PA, Neutral): Key player in the field of Costa, European Capital Goods analyst Smart Factories

Company profile Schneider is a specialist in energy management, offering integrated solutions across multiple market segments, including energy and infrastructure, industrial processes, building automation, and data networks, as well as a broad presence in residential applications. The company operates in more than 100 countries, with c.152,000 employees. The key end markets that Schneider sells into include construction, technology, utilities, oil & gas, mining and general industrial. Of its products, 50% are sold through distributors.

Exposure to the Factory of the Future Schneider’s offering includes software and hardware, which allows for automatic diagnosis, as well as optimisation and configuration of industrial processes through the use of sensors and control devices. Schneider has strengthened its offering here through the acquisition of Invensys in 2013.The Wonderware systems platform is one example of Schneider’s offering in the software space; this is a platform that integrates and consolidates data across the different factory floor systems to be able to function pro- actively rather than reactively. Other products include PlantStruxure, an information and control system for factories, and Machine Struxure, which provides one single software environment with end-to-end service support. Machine Struxure has seen sales grow 3.4x over 2008-13 and connected 200,000 machines in 2013 vs. just 1,000 in 2008. In addition, Schneider also aims to provide a leading digital experience through initiatives such as a 24x7 sales & marketing channel and a tailored customer care experience.

Covered by Jacqueline Siasun Robot (300024.SZ, Neutral): Engineering China’s intelligent Du, China Machinery manufacturing, leading China’s development of cobots & AGVs analyst Company profile Siasun is the largest Chinese robotics automation company, backed by the Chinese Academy of Science. The company was officially founded in 2000, with an initial focus on producing industrial robots for China’s fast-growing automobile market. Currently, it has a complete range of automation-related businesses, including warehouse automation, automatic assembly and testing line, industrial robotics, transportation automation and core robotics parts. Siasun has strong R&D capabilities and has received a number of national awards for robotics, as well as automated guided vehicles. Siasun has partnered with automakers such as Ford Motor, SAIC, Brilliance, Mazda and Dongfeng Motor by providing welding/painting and other robotics. It has also supplied AGVs (automated guided vehicles) to companies such as Shanghai General Motor. Siasun also supplies robots to other industries, such as electronics, textiles and tobacco.

Goldman Sachs Global Investment Research 61 April 13, 2016 Profiles in Innovation

Exposure to the Factory of the Future Siasun is a leading player in China’s factory automation upgrade. The company has reinforced its leading position by expanding its business scale, while constructively deploying its resources to realise its long-term plan of creating a “digital factory”. Siasun is the leader in China’s development of cobots and AGVs, introducing its 7DOF cooperating robots in mid-2015, one year after ABB and Kuka’s offering, and advancing the progress of technology in China towards more flexibility with enhanced sensory capabilities. In the AGV field, Siasun has an approximately 17% market share in China.

Covered by Daniela Siemens (SIEGn.DE, Buy): Strong presence in the industrial Costa, European software space including product lifecycle management Capital Goods analyst Company profile Siemens is Europe’s largest industrial conglomerate, with core activities in the fields of electrification, automation and digitisation and activities in nearly all countries in the world. As of September 2015, Siemens had 348,000 employees. The key end markets that Siemens sells into include utilities, oil & gas, mining, healthcare, rail, construction and general industrial.

Exposure to the Factory of the Future Siemens has bundled its Factory of the Future offerings under the Digital Factory business segment, which provides industry software together with a comprehensive FoF product and service portfolio. Siemens’ product offering aims to transform traditional companies into digital enterprises, increasing the flexibility and efficiency of its production processes and reducing the time to market. The Digital Factory division mainly sells into the manufacturing industry, with the key markets being automotive, aerospace, machine tools and plant installations. Siemens employs 17,500 software engineers and is continually investing in technology and experts in smart data analytics and the cloud. Siemens also has strong IT partnerships with companies such as Atos, Accenture and IBM. A key offering by Siemens in this space is the Product Lifecycle Management software, which provides simulation capabilities for different plant environments and the product lifecycle.

Covered by Yuichiro SMC (6273.T, Neutral): No.1 supplier of the basic automation Isayama, Japan equipment, pneumatics Machinery analyst Company profile SMC was established in 1959 as Sintered Metal Corporation to make and sell sintered metal filters. The company began making pneumatic equipment in 1961 and automatic control equipment in 1964. Today, these are SMC’s core businesses. SMC is a leading maker of pneumatic equipment and has a high market share in many regions: Japan 65%, Europe 20%, the Americas 22%, China 46% and global 33%. Notable features include an operating margin of almost 30% and an inventory strategy that leverages an extremely strong balance sheet (SMC has an inventory turnover period of more than 10-12 months versus a global machinery sector average of three months).

Exposure to the Factory of the Future Introducing pneumatic equipment is probably the easiest way to automate production processes. Pneumatic equipment is not as precise as electric machinery and not as powerful as motor-driven and hydraulic equipment, but in terms of growth potential it has the advantages of being relatively inexpensive and easy to install. Regardless of the introduction of other promising/destructive solutions, as long as there is a need for equipment that converts some kind of signal into output, we forecast pneumatic equipment will benefit from structural demand growth accompanying the advance of automation.

Goldman Sachs Global Investment Research 62 April 13, 2016 Profiles in Innovation

Covered by Sam Eisner, Stratasys (SSYS, Neutral): Leader in prototyping and materials US SMID Capital Goods benefiting from faster 3D printing cycle times analyst Company profile SSYS is a leading manufacturer of rapid prototyping and additive manufacturing products. The company’s product line covers basic desktop models primarily used in prototyping, as well as higher-end professional grade printers capable of producing end use parts. At peak, we believe SSYS’ share of the total 3DP market was roughly 39%, although given recent declines in its low-end platform, this share could be lower today. Key printer technologies include Fused Deposition Modeling and PolyJet. The company provides the “printers” and the materials that they use to make products, as well as service bureaus that create products for customers without the need to purchase a printer. The company has 2,900 employees and operates globally, with principal operations in the US, Israel, Germany and Hong Kong.

Exposure to Factory of the Future Stratasys’ products allow for rapid production, as well as production in small batches, key to make the Factory of the Future more flexible. At the design level, time to market is sharply reduced as iteration on prototypes can occur faster (build times in less than a day vs. multiple weeks for traditional prototyping). On the service side, SSYS is attempting to integrate its service platform with a vertical-based go-to-market model, which attempts to tailor the manufacturing and design process to each application. Lastly, material complexity and cost remain the largest barriers to broader adoption of additive manufacturing. However, SSYS offers 138 “digital materials” that can be coupled with more than 1,000 different colour variations – we see SSYS’s materials leadership, as well as its larger installed base, as key competitive advantages.

Covered by Toshiya Teradyne (TER, Neutral): Market leader in semiconductor testing at Hari, US the head of industry robotics push Semiconductors analyst Company profile Teradyne is a market leader in the highly consolidated System on a Chip test sector (the top 2 participants have nearly 95% share). Teradyne’s Automatic Test Equipment is used to test a variety of products (such as semiconductors, wireless and data storage) employed across a wide spectrum of end markets. The company has four core business units: Semiconductor Test (i.e. logic, RF, analog, power and memory testing), Systems Test (complex electronic systems test), Litepoint (wireless test business), and Collaborative Robots (headed by the Universal Robots division). The company employs 4,100 people worldwide.

Exposure to the Factory of the Future In 2015, Teradyne acquired Universal Robots (UR), a Denmark-based leader in collaborative robotics, and has since built out its collaborative robots division around the acquisition. UR was founded in 2005 and has expanded rapidly, launching its US office in 2012 (it now has offices worldwide). Unlike past robots, which have historically faced a bottleneck at the programming level, UR products are marked by ease of training and programming. According to management, UR is by far an industry leader (it estimates a c.60% market share), and Teradyne expects the market to grow 50%+ over the coming years. Overall, Teradyne views UR products as well positioned to rapidly proliferate across simple manufacturing tasks that can be quickly replicated, and noted that the robots are already more cost-effective than employees in DMs, and also now China. By the end of 2014, there were 3,500 UR robots installed worldwide.

Goldman Sachs Global Investment Research 63 April 13, 2016 Profiles in Innovation

Covered by Yuichiro Yaskawa Electric (6506.T, CL-Buy): Top supplier of servos, and the Isayama, Japan second-largest robotics supplier in the world Machinery analyst

Company profile Founded in 1915, Yaskawa has the top share of the global servo motor market at around 30%, and we estimate its share of the global industrial robot market is around 11%, the second largest behind Fanuc (17%). Yaskawa Electric has a diverse product lineup (in terms of load capacity) that ranges from medium and large robots used in auto production to small vertically articulated robots used in the assembly and inspection of electronic components.

Exposure to the Factory of the Future Yaskawa Electric is developing cobots, and in December 2015, it exhibited a 10-kg load capacity robot at the International Robot Exhibition 2015. Currently, Universal Robot, Rethink Robotics, and other specialist companies are the most well-known makers of cobots. However, we expect Yaskawa Electric, Fanuc, ABB, Kuka, and other major industrial robot makers to make a fully-fledged entry into the cobot field. As is the case with conventional industrial robots, new systems must be introduced with cobots. Yaskawa Electric has accumulated system introduction know-how through its industrial robot business and we believe it will be able to leverage this know-how in the cobot space.

Zebra Technologies (ZBRA, Not Covered): RFID technology specialist with a global footprint and impressive customer base

Company profile Zebra is a global leader in the Automatic Identification and Data Capture (AIDC) market, which consists of mobile computing, data capture, radio frequency identification devices (RFID), barcode printing and other automation products and services. Zebra has sales of more than US$3.6 bn in 2015 and employs approximately 7,000 people. Zebra’s solutions help customers and end-users achieve their strategic business objectives, including improved efficiency and workflow management, increased productivity and asset utilisation, real-time, actionable enterprise information, and better customer experiences.

Exposure to the Factory of the Future Zebra is a pure-play designer, manufacturer, and vendor of a broad range of RFID products, as well as other AIDC technologies including mobile computers, barcode scanners, wireless LAN products, specialty printers for barcode labelling and personal identification, real-time location systems, related accessories and supplies, such as self-adhesive labels and other consumables, and utilities and application software. Manufacturing represents 19% of Zebra’s total sales; the company is one of the largest publically listed RFID specialists, with a global, blue-chip customer base (e.g. Bosch, Ford, Sony, Caterpillar and Boeing).

Goldman Sachs Global Investment Research 64 April 13, 2016 Profiles in Innovation

Drivers & barriers

 Rising labour costs, global competition and customers’ demand for customisation and immediate access to products are the major forces behind the latest FoF technologies being adopted.  Still, lack of standards, young asset bases and social pressure from potential unemployment make the road to the optimal manufacturing landscape potentially longer than technological developments alone would allow.

Goldman Sachs Global Investment Research 65 April 13, 2016 Profiles in Innovation

Ten reasons why we are talking about the ‘Factory of the Future’

As growth slows post a 15-year super-cycle, all eyes are on the manufacturing sector’s margins and returns; we expect these to stagnate over the coming years as operating leverage fades and fixed costs continue to rise (especially labour). Unsurprisingly, the problems have been exacerbated by increasingly unproductive assets, partly owing to legacy assets in DMs and to overinvestment in EMs. Manufacturers will have to increasingly focus on fighting labour cost appreciation in a world where specialised manufacturing talent is becoming scarcer, despite increasingly demanding customers and pressure on companies to return more cash to shareholders. The good news is that more efficient manufacturing practices are now being treated by governments as a priority for nations to stay competitive, and unprecedented developments have taken place in sensing capabilities (the key enabling technology for the Factory of the Future).

Exhibit 52: Slowing growth in manufacturing… Exhibit 53: …has put returns under pressure Average of >800 GS covered manufacturing companies’ sales Average of >800 GS manufacturing companies’ CROCI growth

15.0% 20% 14.5% 15% 14.0% 13.5% 10% 13.0%

5% 12.5% 12.0% 0% 11.5% 11.0% -5% 10.5% -10% 10.0% 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Source: Company data, Goldman Sachs Global Investment Research Source: Company data, Goldman Sachs Global Investment Research

1. DMs need to bring down costs to remain competitive vs. EMs… The average DM manufacturing hourly labour cost is still more than three times higher than the average manufacturing wage in China. The wave of globalisation over the past three decades has seen a shift of manufacturing to low-cost centres in EM as DM labour costs kept rising. More than half of global fixed investment is now in EMs and the growing threat of EM competition, which is climbing up the value chain, is very real for a number of western incumbents (for example, Chinese rail manufacturer CRRC is now more than 3x bigger than the largest western manufacturer). In the face of continuing labour cost increases, DM manufacturers must adopt the latest technologies and enhance productivity to remain competitive.

2. …but rising EM wages are reducing the ‘low-cost location’ edge The by-product of economic development and wealth creation has been EM labour cost inflation far outstripping that of alternative manufacturing locations. Although the cost advantage is still significant, we expect it to continue to narrow as wages rise. In order to maintain or increase their share of global manufacturing, EM factories need to evolve to remain competitive. Several are actively targeting higher efficiency (e.g. Foxconn, one of China’s largest employers, which now designs, manufactures and deploys its own robots has stated its desire to have a “robot army”).

Goldman Sachs Global Investment Research 66 April 13, 2016 Profiles in Innovation

Exhibit 54: Labour costs in DMs significantly exceed EMs... Exhibit 55: …but in EMs, the cost advantage is quickly Annual compensation for a mechanical engineer narrowing 5-year CAGR of annual manufacturing wages in local currency

$90,000 15.0% $81,137 13.0% $80,000 11.0% $70,000 $66,890 9.0%

$60,000 7.0% $49,816 5.0% $50,000 $42,570 3.0% $38,906 $40,000 1.0%

$30,000 -1.0% UK US

$23,056 Italy China 5yr CAGR5yr manufacturing annual wages Japan India * India Russia France Taiwan Norway Average Mechanical Engineer Salary (USD) Salary Engineer Mechanical Average Canada Mexico* Belgium $20,000 Sweden $15,707 Estonia* Australia Germany Singapore

$10,000 $5,387 South Korea ● Developed market * Indicates where manufacturing wages were not available and average total $0,000 ● Emerging market Switzerland US Germany UK France China Mexico India wages were used instead.

Source: PayScale (January 2016). Source: International Labor Organization, Haver, Trading Economics, Eurostat.

3. Specialised manufacturing labour is increasingly scarce Not only rising labour costs are forcing higher adoption of technologies to streamline manufacturing costs; there is also a growing shortage of skilled manufacturing labour, as noted by 84% of executives interviewed in the US recently by the Manufacturing Institute. They noted that six out of ten open skilled production positions are unfilled owing to a shortage of talent. The scarcity of specialised labour is exacerbated in the largest manufacturing nations by ageing populations and in EMs by a reduced desire among younger generations to work in factories. By 2030, McKinsey expects a shortage of both highly skilled workers (engineers, technicians) and medium-skilled workers (technicians, factory workers) in Brazil, China and India – driven by the rapid growth in knowledge- intensive manufacturing.

Exhibit 56: The labour force is ageing… Exhibit 57: … and skills required are changing Average age of population Survey of 250 industry executives regarding the future of talent (2015)

50 100% 6% 5% 90% 80% 45 36% 70% 52%

Japan 60% 40 Germany 50% S.Korea 40% 30% China 55% 35 France 20% 42% UK 10%

Average population age population Average USA 0% 30 New education and training will be required to The increase use of automation will transform meet the demands of the future digital job the skill mix in the future workforce market

25 Strongly agree Agree Neither agree nor disagree Disgree Strongly disagree I don't know 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016E 2018E 2020E 2022E 2024E 2026E 2028E 2030E 2032E 2034E

Source: UN, Goldman Sachs Global Investment Research . Source: World Economic Forum .

Goldman Sachs Global Investment Research 67 April 13, 2016 Profiles in Innovation

4. Productivity is a differentiator in a world with legacy capacity The past 15 years saw an unprecedented level of fixed asset investment globally. This was driven primarily by material investment in Chinese infrastructure, and we do not expect it to be repeated over the next decade (see our report Capex Tracker: No end in sight to the capex slide, September 18, 2015). Simultaneously, this investment wave drove a new set of low-cost competitors to emerge in key capital goods areas such as electrical equipment, rail and chemicals production. This has led to overcapacity across many global industries, including general industrial manufacturing. One of the key benefits of the upcoming manufacturing technology profiled in this report is not just that it can bring down costs, but that it can increase the total potential output per unit of equipment, allowing companies to improve returns by cutting capacity, but still fulfilling demand.

Furthermore, capital goods companies that serve manufacturing need to move in two interlinked ways: (1) increasingly penetrate their installed bases; and (2) find new services/products to differentiate their offer vs. old and new competitors. The “digitisation” of many products and services is in the sweet spot for this.

Exhibit 58: Productivity is at a trough… Exhibit 59: …and manufacturing capacity utilisation Market cap-weighted asset turn (sales/GCI) for GS global remains low… coverage ex-financials US and Europe manufacturing capacity utilisation

1.05x 90%

1.00x 85%

0.95x 80%

0.90x 75%

0.85x 70% 0.80x 65% 0.75x 60% 0.70x

Coverage asset turn (MC weighted) (MC turn asset Coverage 55% 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016E 2017E 2018E 1985 1986 1987 1988 1990 1991 1992 1993 1995 1996 1997 1998 2000 2001 2002 2003 2005 2006 2007 2008 2010 2011 2012 2013 2015 Market cap weighted asset turn (sales/GCI), GS analyst forecast for global coverage (ex- financials) Europe Manufacturing Capacity Utilization US Manufacturing Capacity Utilization

Source: Company data, Goldman Sachs Global Investment Research. Source: Federal Reserve, Eurostat.

Exhibit 60: …in DM owing to legacy assets… Exhibit 61: ...and in EMs owing to overinvestment Average age of manufacturing assets in the US; 1947-2014 Fixed investment as a % of GDP; global and China

10 27% 50%

9.5 45% 26% 9 40% 25% 35% 8.5 30% 8 24% 25% 7.5 23% 20% 7 22% 15% 6.5 Age of equipment in US in the of equipment Age

Fixed Investment % of GDP of % InvestmentFixed 10% GDP of % InvestmentFixed 6 21% 5%

1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 20% 0% Manufacturing Manufacturing median 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Global China (RHS)

Source: BEA Source: Haver, Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 68 April 13, 2016 Profiles in Innovation

5. There is an increased push to shorten time to market, particularly by eliminating stranded inventories Lengthy product development times are costly and manufacturers are rapidly shortening their time to market. Maserati, an automotive brand of Fiat Chrysler, has reduced time-to- market to 16 months from 30 months by introducing some of the latest FoF technologies. Today’s market environment means information comes faster and is more “Industrie 4.0 basically takes the cost of accessible than ever, and that customers scale close to zero. No matter what lot now expect products sooner. size you need, the unit cost is about the same. At some point, what will happen is Faster time to market also means a lower this: You are a consumer and you want gap from order to delivery on a day-to-day to buy a car. You go to the Internet, put basis. This is a key source of being able to your specs together, and send that order generate better return on capital for a to BMW. Someone will check your credit manufacturer, especially when available history and your funds. Then, your car capacity is not scarce. Technologies that will go straight to production and the can integrate communication between factory will build it to order. Four weeks suppliers, manufacturers, distributors and from now, you will have a car. No more waiting six months or compromising at the final customers are a powerful tool to the dealership, where they have 50 cars eliminate stranded inventories. Atos, an IT but not the one you want.” services provider, estimates that manufacturing IoT platforms can reduce Joe Kaeser, CEO of Siemens, on the future of idle time by up to 58%.

Exhibit 62: Overall inventory levels have not materially Exhibit 63: Demand is rising for shorter time to market declined outside Europe over the past years Survey of 100 companies (2015) Inventory days across GS manufacturing coverage by region

115 114.5 Increased demand for shorter time to market 110 60%

105 50% 102.1 100 40% 95

90 30%

85 82.7 20% inventory indexed Days to 2009 level 80 10% 75

0% 2009 2010 2011 2012 2013 2014

2015E Disagree Somewhat Somewhat agree Agree Don't know disagree China Europe US

Source: Company data, Goldman Sachs Global Investment Research. Source: 3Gamma

6. Customers demand unprecedented customisation Personalisation today is used as a competitive tool to capture sales, and is something that customers look for to distinguish their purchases. In addition, the growth of consumers from emerging markets, encompassing a diverse range of cultural and ethnic groups, increases the complexity of manufacturing in order to appeal to these new large markets. Finally, manufacturers must also prepare for further proliferation of models and greater customisation driven by shorter life cycles.

Goldman Sachs Global Investment Research 69 April 13, 2016 Profiles in Innovation

Exhibit 64: Customers are demanding personalised products Combinations available for a customer buying a Ford F150 pickup

Theoretical Equipment options Variants combinations Trim 6 6 Passenger compartment 3 18 Power train 2 36 Cargo space 4 144 Engine 3 432 Transmission 3 1,296 Rear axle ratio 7 9,072 Wheels 9 81,648 Tires 8 653,184 Seats 18 11,757,312 Power seats 2 23,514,624 Radio 5 117,573,120 Running boards 4 470,292,480 Rear windows 3 1,410,877,440 Colors 12 16,930,529,280 Interior trim colors 3 50,791,587,840 16 individual options 12,870 653,687,735,500,800

Source: Siemens.

7. Focus on safety and security has increased significantly… The emphasis on worker and product safety has scaled up significantly over the past few years. A more automated, controlled and less labour-intensive environment will reduce the likelihood of accidents and costly litigation. In addition, factories have to be safe from cyber-attacks given the increasingly important role data plays in the manufacturing process.

Exhibit 65: High costs of safety incidents… Exhibit 66: …and increased focus on security Cost to society of work injuries in the UK (£ in 2013 prices) Survey of 3,382 IT executives in Asia-Pacific; % answering “increase 5-10% or more”

Human Financial Total cost "How do you expect your spending on the following software categories cost cost to change in 2014 compared with 2013?"

Fatal injuries £1,153,000 £421,600 £1,574,600 Security software 55% Non-fatal Infrastructure software 46% £4,600 £2,900 £7,500 injuries Business intelligence 46% 7 or more days SaaS (collaboration and email) 41% £17,600 £10,100 £27,700 absence Collaboration application 40% Up to 6 days Big data solutions 40% £330 £550 £880 absence Platform software 39% Packaged process applications Ill health £9,900 £8,700 £18,600 38% SaaS (CRM and HR Process) 38% 7 or more days £20,100 £17,300 £37,400 SaaS (industry specific process) 37% absence Mobile applications and mobile middleware 17% Up to 6 days £290 £560 £850 absence

Source: HSE. Source: Forrester.

8. …and some governments are actively pushing to stay ahead in manufacturing In a globalised world with lower trade barriers, scarcer demand and more supply, different countries have to fight hard for competitiveness to be suppliers of choice for manufacturing. We have seen numerous initiatives over the past few years that stress this.

For example, Germany has introduced the Industrie 4.0 initiative, representing the fourth industrial revolution, through which it is promoting the concept of ‘smart factories’. The use of cyber-physical systems and the communications between components allow for less centralised process management, providing significant quality, time, resource, and cost advantages in comparison with traditional production systems. Industrie 4.0 was a term that originated in industry, but the initiative has received backing by the German government and is a key feature of its High-Tech Strategy 2020 Action Plan.

Goldman Sachs Global Investment Research 70 April 13, 2016 Profiles in Innovation

“China Manufacturing 2025” is another example of a government initiative to advance national manufacturing industries. The objective is to move up the value chain, highlighting ten specific industries, centering on five projects: (1) build national/provincial R&D centres; (2) promote intelligent manufacturing pilot programmes; (3) fix the weak link between components, crafts and materials; (4) promote green manufacturing and energy saving; and (5) promote high-end equipment innovation.

In the US, there are several initiatives to promote advanced manufacturing, including the Smart Manufacturing Leadership Coalition, the Industrial Internet Consortium and the Advanced Manufacturing Partnership.

9. Short-term demands from shareholders for dividends and buybacks puts further strain on cash available for organic growth Short-term macro uncertainty and a lack of visibility on growth have led to increasing demand from investors for cash returns from corporates (dividends and buybacks). While this might not be a sustainable means of capital allocation in the long term, it has weighed on short-term considerations for corporates and put further pressure on their ability to optimise FCF generation through more productive capex.

Exhibit 67: Investors prefer dividends/buybacks… Exhibit 68: …and companies are listening to them GS European Industrials survey of 57 investors OCF used to invest in growth vs. returning to shareholders

85% 36% 100% Spend more on 80% 90% 24% restructuring of existing 33% 80% operations 75% 30% 70% Step up M&A 70% 19% 60% 65% 27% 50% 60% 17% Raise capex 24% 40% 55% 21% 30% 50% Return cash to 20% 40% shareholders 45% 18% 10% 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

0% 2016E 2017E 2018E March 2016 Invest for Growth (capex + R&D + M&A) (LHS) Return to Investors (buybacks + dividends) (RHS)

Source: Goldman Sachs Global Investment Research Source: Company data, Goldman Sachs Global Investment Research

10. Finally (and most importantly), key technologies now exist for fully optimised and connected manufacturing Rapid improvements have taken place in the capabilities of a number of technologies, which have the potential to drive substantial change in factories across the world. While the cost of some technologies has fallen sharply, partly due to the smartphone revolution, widespread adoption is becoming increasingly likely. For example, processing costs have declined nearly 60x over the past ten years, enabling more devices to analyse data. Sensing technology development in particular has meant that many manufacturing industries are at the start of a period of unprecedented change. According to McKinsey, the price of microelectromechanical sensors has declined by 80%-90% over the past five years. Sensors are the key bit of hardware that bridges the digital world with the real world, and declines in sensor prices have also spearheaded declines in the prices of other manufacturing technologies, such as robotics – with claims by some industry participants that prices have declined by up to 10% pa in recent years.

Goldman Sachs Global Investment Research 71 April 13, 2016 Profiles in Innovation

Exhibit 69: In general, the cost of sensors has more than halved over the past decade, with the decline being even more pronounced for specific sensor applications in manufacturing

$1.40 $1.30 $1.20 $1.11 $1.00 $0.95

$0.80 $0.82 $0.70

$0.60 $0.60 $0.51 $0.44 $0.40 $0.38

$0.20 2004 2006 2008 2010 2012 2014 2016E 2018E 2020E

Average sensor cost

Source: Goldman Sachs Global Investment Research, BI Intelligence.

Venture capital investment in robotics and software is heating up Venture capital and private equity investments in robotics have picked up in recent years, and manufacturing, warehousing and industrial software investments have also risen significantly, albeit from a low base. Unsurprisingly, given the maturity of RFID technologies, VC/PE investment remains stable. However, this makes up only a small fraction of the US$400 bn+ in venture capital and private equity deals completed in 2014 (where more than 70% were in software-related businesses, healthcare and telecoms).

Exhibit 70: Investment in robotics and software is heating up 12m rolling VC/PE investment in the US (US$ mn)

450

400

350

300

250

200

150

100

12m rolling VC/PE invest. in in US ($mn) invest. VC/PE rolling 12m 50

0 12 Q1 12 Q2 12 Q3 12 Q4 13 Q1 13 Q2 13 Q3 13 Q4 14 Q1 14 Q2 14 Q3 14 Q4 15 Q1 15 Q2 15 Q3 15 Q4 16 Q1 Robotics RFID Manufacturing, warehousing & industrial software

Source: CBInsights, Goldman Sachs Global Investment Research

Goldman Sachs Global Investment Research 72 April 13, 2016 Profiles in Innovation

What could derail it?

Manufacturing is entering a decade of significant change and we have tackled some of the latest technology developments. However, predicting the future of manufacturing comes with innate challenges, as no-one can predict the full impact of technological innovation. Technologies are adopted at unpredictable rates and entirely unexpected applications can arise and dominate their uses. Below, we highlight the potential barriers to adoption that could derail the evolution of manufacturing:

 Lack of one single protocol and standard has been a recurring concern for a number of suppliers and customers. This includes, for example, a commonly understood programmable robotics language and allowing compatibility/integration between the machines of various suppliers. For example, looking back at the electronics industry, PCs only became widely adopted once the systems were standardised and user-friendly.

 Lack of incentives owing to excess capacity in manufacturing industries could prevent adoption of/investment in new technologies. We highlight that while capacity may be not be stretched, pressure on margins is likely to drive disruption in manufacturing in order to increase operating leverage.

 Limited cost deflation: There is currently a clear deflationary cost trajectory for many of the key technologies, but there is no certainty that this will continue.

 Potential effect on employment may generate social and political unrest. Ultimately, one of the key incentives to automate is to reduce wage costs. In China, for example, over 230 mn people are employed in manufacturing jobs: the social repercussions could be significant if unemployment were to rise rapidly should the labour force not have time to adapt and find alternative work.

 Longevity of capital goods assets postponing adoption. The rapid progress of technological innovation in manufacturing could be delayed by the longevity of the capital goods inputs (the average industrial robot lasts 12-15 years). If the benefits of employing new equipment do not strongly outweigh the total costs incurred (including installation costs, prior arrangement redundancy costs and fixed infrastructure adjustments), investment may be delayed.

 Decreases in demand for traditional industrial capex products as we shift towards more opex-driven business models (e.g. equipment as a service). Firms may not wish to make the heavy investment needed to automate their factories. We note automation capex will ultimately reduce opex.

 IP issues could prevent the Internet of Things from reaching its full potential. The question of “who does the data belong to?” may prevent data from being shared between machines owing to the potential proprietary and security-sensitive information it may contain. Our discussions with companies suggest that data generated from equipment consensually belongs to the client.

 Development costs might be too high for equipment makers to pursue revolutionary new products. As Google’s recent announcement of the sale of its robotics arm Boston Dynamics suggests, corporates might not be willing to take the burden of multiple years of development costs for products with a lack of prospects for meaningful near-term revenues. This is exacerbated by the threat of slow adoption given relatively new legacy installed bases (2000-13 saw substantial capacity investments) and a conservative set of customers (compared with the less risk-averse and ROIC-sensitive consumer segment).

Goldman Sachs Global Investment Research 73 April 13, 2016 Profiles in Innovation

 Overautomation. In addition to a lack of system integration skills and bad selection of FA equipment, we think overautomation is the common cause of bad automation.

• Additional overhead costs from over-automation may outweigh the savings from direct labour, in particular when capacity utilisation falls. As factories increase the pace of automation, overhead costs grow and direct labour costs fall. Increased support costs associated with maintaining and running automated equipment drive overhead costs up even more. The increase in operational leverage owing to a higher proportion of fixed costs hits particularly hard when capacity utilisation falls in a downturn.

• Overautomation with the wrong equipment reduces flexibility. While some of the technologies highlighted in this report combat this challenge, most of today’s machines/robots are good at performing one job at very high efficiency levels but with limited flexibility. This is a common argument for not automating. Take a robot gripper as an example. If it only has two positions, the gripper can only grab parts that it has been designed for. If it is to grab different products, the gripper needs to be more complex, more expensive, and often less effective.

Exhibit 71: Overautomation reduces flexibility Performance vs. flexibility of a robot gripper

Source: Goldman Sachs Global Investment Research.

Automation gone wrong: Increasing overheads and decreasing flexibility

According to a paper by the World Congress of Engineering on Volkswagen’s automation strategy, for certain sections of the Golf A5 production line in Germany, it is optimal to limit automation as the additional personnel needed for maintenance already offsets the saved labour cost, and readily available qualified workers produce similar quality while offering much better flexibility. In terms of productivity, the paper finds that in certain sections well-trained manual workers produce a higher percentage of faultless parts, as highly complicated automation systems are susceptible to faults. Even in EMs such as China, over-automation may harm productivity as a result of inexperienced domestic system integrators and an absence of qualified maintenance personnel for the automation equipment. A Chinese local news report stated that in Zhejiang province, where a “robot for human” shift was greatly encouraged, some companies had to set their Japan-made industrial robots aside for two weeks while waiting for the Japanese engineers to arrive to fix technical issues that had arisen.

Goldman Sachs Global Investment Research 74

April 13, 2016April 13, Goldman Sachs Global Investment Research Appendix 1: Key names in the Factory of the Future competitive landscape

Exhibit 72: Key names in the competitive landscape (A-H) Non-exhaustive list of public and private companies

Manufacturing design & Physical Maunfacturing In-factory logistics Other technologies production simulation Additive Collaborative Augmented Machine Nano- Demand Other auto- Company Type Location Ticker IoT PaaS PLM software AGVs RFIDs Machine vision manufacturing robots reality learning technology response mation equip. 1 3D Systems Public US DDD  2 ABB Public Switzerland ABBN.VX   3 Advantech Public Taiwan 2395.TW  4 Alien Technologies Private US n.a.  5 Alphabet (Google) Public US GOOGL  6Amazon Public US AMZN  7 Ansys Public US ANSS  8 Arena Solutions Private US n.a.  9 Atos Public France ATO.PA  10 Autodesk Public US ADSK  11 Automata Technologies Private UK n.a.  12 Aveva Public UK AVV.L  13 Beckhoff Private Germany n.a.  14 Bosch Private Germany n.a.   15 Bruker Public US BRKR  16 CipherLab Public Taiwan 6160.TWO  17 Cisco (Jasper) Public US CSCO  18 Clearpath Private US n.a.  19 Cognex Public US CGNX  20 Cumulocity Private Germany n.a.  21 Daifuku Public Japan 6383  22 Dassault Systems Public France DAST.PA  23 Delta Electronics Public Taiwan 2308.TW  24 Dematic Private US n.a.  25 Device Insight Private Germany n.a.  26 Duerr Public Germany DEUG.DE  27 Duetsche Post (DHL) Public Germany DPW.DE  28 Efort Private China n.a.  29 EOS Private Germany n.a.  30 Epson Public Japan 6724.T  31 Estun Public China 002747.SZ  32 Fanuc Public Japan 6954.T  33 FedEx Public US FDX  34 Forcam Private Germany n.a.  35 GE Public US GE    36 GSK CNC Equipment Private China n.a.  37 Han's Laser Public China 002008.SZ    38 Harmonic Drive Systems Public Japan 6324.T  39 Hewlett-Packard Public US HPQ  40 Hexagon Public Sweeden HEXAb.ST  41 HG Tech Public China 000998.SZ  Profiles in Innovation 42 Hiwin Public Taiwan 2049.TW  43 Hollysys Public China HOLI  44 Hon Hai (Foxconn) Public Taiwan 2317.TW  75 45 Honeywell Public US HON 

Source: Goldman Sachs Global Investment Research.

April 13, 2016April 13, Goldman Sachs Global Investment Research Exhibit 73: Key names in the competitive landscape (I-Z) Non-exhaustive list of public and private companies

Manufacturing design & Physical Maunfacturing In-factory logistics Other technologies production simulation Additive Collaborative Augmented Machine Nano- Demand Other auto- Company Type Location Ticker IoT PaaS PLM software AGVs RFIDs Machine vision manufacturing robots reality learning technology response mation equip. 46 IBM Public US IBM   47 Impinj Private US n.a.  48 Ingenu Private US n.a.  49 INSYS Microelectronics Private Germany n.a.  50 JBT Corporation Public US JBT  51 Junghenirich Public Germany JUN3.DE  52 Kawada Public Japan 3443.T  53 Kawasaki Robots Public Japan 3045.T   54 Keyence Public Japan 6861.T  55 KION Public Germany KGX.DE  56 KUKA Public Germany KU2G.DE   57 LifeRobotics Private Japan n.a.  58 Microsoft Public US MSFT    59 MioSoft Private US n.a.  60 Misumi Public Japan 9963.T  61 MJ Intelligent System Private China n.a.   62 Nidec Public Japan 6594.T  63 Omron Public Japan 6645.T  64 Oracle Publc US ORCL  65 Oxford Instruments Public UK OXIG.L  66 PTC Public US PTC   67 Renishaw Public UK RSW.L  68 Rethink Robotics Private US n.a.  69 Rockwell Automation Public US ROK   70 Salesforce Public US CRM  71 SAP Public Germany SAP.DE  72 Savant Automation Private US n.a.  73 Schneider Public France SCHN.PA   74 Seegrid Private US n.a.  75 Shenzhen Inovance Technology Public China 300124.SZ  76 Siasun Public China 300024.SZ    77 Siemens Public Germany SIEGn.DE    78 Smartrac Public Netherlands SM7v.DE  79 SMC Public Japan 6273.T  80 Spectris Public UK SXS.L  81 STEP Public China 002527.SZ  82 Stratasys Public US SSYS  83 THK Public Japan 6481.T 84 Tixi Private Germany n.a.  85 Toshiba Public Japan 6502.T  86 Toyota Motors Public Japan 7203.T  87 Teradyne (Universal Robots) Public US TER  88 Trumpf Private Germany n.a.   89 Yaskawa Public Japan 6506.T   90 Yokogowa Public Japan 6841.T   Profiles in Innovation 91 Zebra Technologies Public US ZBRA  

76 Source: Goldman Sachs Global Investment Research

April 13, 2016 Profiles in Innovation

Appendix 2: Toyota Production System and Kanban manufacturing

IoT ecosystems are not entirely new: Toyota Production System and Kanban manufacturing

Toyota Production System (TPS) already has many similarities with the emerging digital manufacturing world. TPS achieves a just-in-time system capable of handling small-lot production of a large variety of products by using a Kanban approach and developing multi-skilled workers. What both IoT enabled and Industrie 4.0 systems have in common is that they allow production to be adjusted to changes in demand by using data linking media (IoT in one case and automation based on multi-skilled workers/smart machines in the other). Both Industrie 4.0 and TPS are good at handling changes in demand because information between factories/processes is quickly shared.

Exhibit 74: Comparison of Industrie 4.0 production flow vs. Toyota Production System production flow

Source: Goldman Sachs Global Investment Research.

Exhibit 75: Comparison of production adjustments between factories/processes

Source: Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 77 April 13, 2016 Profiles in Innovation

Appendix 3: Glossary of terms

Exhibit 76: Glossary of terms

Term Definition

3D Printing or Additive Manufacturing; is the formation of 3D solid objects from a digital motel, typically by laying thin layers of a material AGV Automated Guided Vehicle ; automatic material handling equipment CAM Computer Aided Manufacturing; the use of software to help model the manufacturing of a product. Cloud Decentralised storage, management and processing of data Cobot Lower-cost, smaller and more flexible robots capable of working alongside humans Everything as a Service; pay-by-usage/subscription-based models for machinery, transferring capex into opex for manufacturers and creating a perpetuation of EAAS revenue streams instead of a one-off asset sale for suppliers. Enterprise Resource Planning; business-management software that can be used to collect, store, manage and interpret data from many business activities, including: ERP product planning, purchase. manufacturing or service delivery. IoT Internet of Things ; the network that connects devices and objects. Sensors and software allow for the collection and exchange of data. IoT PaaS Internet of Things Platform as a Service; Industrial software as a service which is based on the cloud that allows machines to communicate and optimise production. M2M Machine-to-Machine; refers to the communication and exchange of data between machines and other other devices Machine learning or artificial intelligence; is the ability of machines to process data into information and derive knowledge from that information to act independently. Technologies with imaging-based automatic inspection, gauging, counting and analysis at high speeds, reliability and with greater precision (exceeding the capabilities Machine vision of the human eye). MES Manufacturing Execution Systems; The central computational system which tracks and documents the transformation of raw materials to finished goods Programmable logic controller; a digital computer used for automation of typically industrial electromechanical processes, such as control of machinery on factory PLC assembly lines, amusement rides, or light fixtures PLM Product Lifecycle Management; using advanced computational methods to create a simulation of a product’s production and its life cycle Radio-frequency Identification devices; technologies that use radio waves to transfer to data for the purposes of automatically identifying and tracking tags attached to RFID objects

Source: Goldman Sachs Global Investment Research.

Goldman Sachs Global Investment Research 78

April 13, 2016April 13, Goldman Sachs Global Investment Research Appendix 4: Price targets, methodologies and key risks

Exhibit 77: Covered companies profiled in the report

Ticker Name Currency Price Rating Price target Key risks Valuation Methodology Upside: Direct Metal Laser Sintering ("DMLS") adoption, killer app discovery; Downside: Timing of new product introductions, DDD 3D Systems Corp. USD 16.63 Neutral 10.00 50/50 Weighting of DCF and 1.2x EV/Sales execution on consumer strategy, integration risk.

ABBN.S ABB Ltd. USD 18.87 Sell 17.00 Upside risks: M&A, accelerated buyback/key shareholders raising stakes, higher earnings. 2017E EV/IC vs ROIC/WACC

End market demand; currency volatility (greatest exposure is to Euro, USD and RMB); higher-/lower-than expected operating 12m EV/GCI vs. CROCI/WACC (cash return multiple at 2395.TW Advantech Corporation TWD 224 Neutral 200 expenses; lower/higher than expected synergies/returns on M&A. 13.8X) Primary risks include macro, service provider capex, commoditization, competition, and execution around the recent executive CSCO Cisco Systems Inc. USD 27.64 Neutral 32 14X our CY16 non-GAAP EPS transitions

DAST.PA Dassault Systemes EUR 69.84 Buy 89.00 M&A integration; sales execution; higher investments; and FX exposure 29x CY17E PF EPS

12m EV/GCI vs. CROCI/WACC (cash return multiple at 2308.TW Delta Electronics TWD 138 Buy 208 Global economic slowdown, lower-than-expected synergies from Eltek acquisition, higher-than-expected operating expenses 13.8X)

6954.T Fanuc JPY 18235 Buy 19500 Greater-than-expected decline in FA demand and weak autos/general industry demand in Europe, US, and China FY3/17-18avg. EV/GCI v CROCI/WACC

Upside: Further capital deployment, greater industrial cost‐outs; Downside: Execution missteps, O&G weakness, softer GE General Electric Co. USD 30.81 Neutral 28.00 P/E (19.0x 2016 EPS) orders/pricing.

6324.T Harmonic Drive Systems Inc. JPY 2976 Buy 3100 Yen appreciation, weaker-than-expected cobot demand, stiffer competition spurred by new entrants FY3/21E EV/GCI vs. CROCI/WACC

2017E EV/GCI vs CROCI/WACC with cash return multiple 002008.SZ Han's Laser Technology Industry Group CNY 22.95 Buy 31.50 Downside risk: uptick in competition from local/foreign peers; less-than-expected smartphone capex of 3.1x with 35% discount

HON Honeywell International Inc. USD 113.03 Buy 126.00 Slower non res/aero/O&G growth; execution missteps P/E (19.0x 2016 EPS)

Downside risk: Further shrinking margins caused by competition or weaker-than-expect cost control; less favorable government 2017E EV/GCI vs CROCI/WACC with cash return multiple 300124.SZ Shenzhen Inovance Technology Co., Ltd CNY 38.39 Buy 50.92 support on NEV industry of 3.1x with 5% premium

6861.T Keyence JPY 61800 Neutral 66000 Forex fluctuations, FA demand trends in all regions, particularly Japan, Europe, the US, and China FY3/17E-FY3/18E average CROCI/WACC vs. EV/GCI,

6594.T Nidec JPY 7437 Buy 10000 Sharper than expected HDD decline, yen appreciation. FY3/17-18avg. EV/GCI v CROCI/WACC

Upside: Acceleration in US/Process; better productivity; accretive M&A; Downside: O&G project delays; weaker pricing; softer ROK Rockwell Automation Inc. USD 112.99 Neutral 101.00 P/E (17.1x CY2016 EPS) industrial backdrop

SAPG.DE SAP EUR 68.45 Buy 95.00 Macro headwinds; lack of traction in HANA or cloud; integration/execution issues with M&A; management changes. 21x 2017E PF EPS

Downside: greater-than-expected weakness in China, deteriorating oil & gas demand, delays with Invensys’ integration, lower SCHN.PA Schneider Electric EUR 53.69 Neutral 57.00 2017E EV/IC vs ROIC/WACC volume/price, FX headwinds, large value-destructive M&A. Upside are the reverse of these. 2017E EV/GCI vs CROCI/WACC with cash return multiple 300024.SZ Siasun Robot&Automation Co., Ltd CNY 27.56 Neutral 22.49 Upside risks: Better-than-expected orders from new industry other than auto/auto parts; Downside risk: local/foreign competition of 3.1x with 70% premium Downside: (1) slower restructuring of underperforming businesses; (2) execution risks; (3) value-dilutive M&A; (4) weaker FCF and SIEGn.DE Siemens AG EUR 89.81 Buy 102.00 2017E EV/IC vs ROIC/WACC volume/price.

6273.T SMC JPY 26805 Neutral 30000 Fluctuations in FA demand, forex FY3/17-18avg. EV/GCI v CROCI/WACC

Upside: Direct Metal Laser Sintering ("DMLS") adoption, killer app discovery; Downside: Timing of new product introductions, SSYS Stratasys Ltd. USD 27.27 Neutral 21.00 50/50 Weighting of DCF and 1.2x EV/Sales execution on consumer strategy, integration risk.

TER Teradyne Inc. USD 20.65 Neutral 22.00 Semi test growth, cycle trajectory, and Universal Robots execution 12m 10x normalised EPS

6506.T Yaskawa Electric JPY 1336 Buy 1560 Yen appreciation, product mix improvement below our expectations, far lower-than-expected FA demand FY3/17-18avg. EV/GCI v CROCI/WACC

Source: Goldman Sachs Global Investment Research (all price targets are 12-months). Profiles in Innovation

79

April 13, 2016 Profiles in Innovation

Disclosure Appendix Reg AC We, Daniela Costa, Yuichiro Isayama, Joe Ritchie, Tian Lu, CFA, William Turner, Ashay Gupta, Yuki Kawanishi, Mohammed Moawalla, Jacqueline Du, Samuel H. Eisner, Willy Chen, Toshiya Hari, Stefan Burgstaller, Daiki Takayama, Simona Jankowski, CFA, Wei Chen, Heather Bellini, CFA, Diana Zhao, Alex Karpos, Matthew Cabral, Gautam Pillai, Jessica Kaur, Shateel Alam and Robert D. Boroujerdi, hereby certify that all of the views expressed in this report accurately reflect our personal views about the subject company or companies and its or their securities. We also certify that no part of our compensation was, is or will be, directly or indirectly, related to the specific recommendations or views expressed in this report.

Unless otherwise stated, the individuals listed on the cover page of this report are analysts in Goldman Sachs' Global Investment Research division. Investment Profile The Goldman Sachs Investment Profile provides investment context for a security by comparing key attributes of that security to its peer group and market. The four key attributes depicted are: growth, returns, multiple and volatility. Growth, returns and multiple are indexed based on composites of several methodologies to determine the stocks percentile ranking within the region's coverage universe. The precise calculation of each metric may vary depending on the fiscal year, industry and region but the standard approach is as follows: Growth is a composite of next year's estimate over current year's estimate, e.g. EPS, EBITDA, Revenue. Return is a year one prospective aggregate of various return on capital measures, e.g. CROCI, ROACE, and ROE. Multiple is a composite of one-year forward valuation ratios, e.g. P/E, dividend yield, EV/FCF, EV/EBITDA, EV/DACF, Price/Book. Volatility is measured as trailing twelve-month volatility adjusted for dividends. Quantum Quantum is Goldman Sachs' proprietary database providing access to detailed financial statement histories, forecasts and ratios. It can be used for in-depth analysis of a single company, or to make comparisons between companies in different sectors and markets. GS SUSTAIN GS SUSTAIN is a global investment strategy aimed at long-term, long-only performance with a low turnover of ideas. The GS SUSTAIN focus list includes leaders our analysis shows to be well positioned to deliver long term outperformance through sustained competitive advantage and superior returns on capital relative to their global industry peers. Leaders are identified based on quantifiable analysis of three aspects of corporate performance: cash return on cash invested, industry positioning and management quality (the effectiveness of companies' management of the environmental, social and governance issues facing their industry). Disclosures Coverage group(s) of stocks by primary analyst(s) Daniela Costa: Europe-Machinery & Elec Equip. Yuichiro Isayama: Japan-Machinery. Joe Ritchie: America-Capital Goods: Multi-Industry. Tian Lu, CFA: Asia Pacific Infrastructure, China Capital Goods. Mohammed Moawalla: Europe-IT Services, Europe-Software. Jacqueline Du: Asia Pacific Infrastructure, China Capital Goods. Samuel H. Eisner: America-Additive Manufacturing, America-Building Products, America-SMID Capital Goods. Willy Chen: Asia Pacific Autos & Auto Parts, Asia Pacific Chemicals, Asia Pacific Energy, Taiwan Machinery. Toshiya Hari: America-Semiconductor Capital Equipment, America-Semiconductors. Stefan Burgstaller: Europe-Autos & Auto Parts. Daiki Takayama: Japan-Electronic Components. Simona Jankowski, CFA: America-Consumer Hardware & Mobility, America-IT Hardware, America-Telecom Equipment. Wei Chen: Asia Pacific Technology. Heather Bellini, CFA: America-Software. Matthew Cabral: America-IT Hardware. America-Additive Manufacturing: 3D Systems Corp., Stratasys Ltd.. America-Building Products: A.O. Smith Corp., Apogee Enterprises Inc., Armstrong World Industries Inc., BMC Stock Holding Inc., Fortune Brands Home & Security Inc., Lennox International Inc., Masco Corp., Mohawk Industries Inc., Owens Corning, Watsco Inc., Whirlpool Corp.. America-Capital Goods: Multi-Industry: 3M Co., Colfax Corp., Dover Corp., Eaton Corp., Emerson Electric Co., Flowserve Corp., General Electric Co., Graco Inc., HD Supply Holdings, Honeywell International Inc., Illinois Tool Works, Ingersoll-Rand Plc, ITT Corp., Parker Hannifin Corp., Pentair Plc, Rockwell Automation Inc., Roper Technologies Inc., Tyco International Plc, W.W. Grainger Inc.. America-Consumer Hardware & Mobility: Apple Inc., BlackBerry Ltd., BlackBerry Ltd., Corning Inc., Garmin Ltd., GoPro Inc., Qualcomm Inc.. America-IT Hardware: Aerohive Networks Inc., Arista Networks Inc., Brocade Communications Systems, CDW Corp., Cisco Systems Inc., EMC Corp., F5 Networks Inc., Hewlett Packard Enterprise Co., HP Inc., Motorola Solutions Inc., NetApp Inc., Nimble Storage Inc., Pure Storage Inc., Ruckus Wireless Inc., Xerox Corp.. America-SMID Capital Goods: Kennametal Inc., Milacron Holdings, RBC Bearings Inc., Regal Beloit Corp., Rexnord Corp., ServiceMaster Global Holdings, Timken Co., TriMas Corp., Wabtec Corp.. America-Semiconductor Capital Equipment: Applied Materials Inc., Keysight Technologies Inc., Lam Research Corp., SunEdison Semiconductor Ltd., Teradyne Inc.. America-Semiconductors: Broadcom Ltd., Intel Corp., NXP Semiconductors NV, Qorvo Inc., Skyworks Solutions Inc., Texas Instruments Inc., Xilinx Corp.. America-Software: Adobe Systems Inc., Akamai Technologies Inc., Alarm.com Holdings, Alphabet Inc., Atlassian Corp., Autodesk Inc., Citrix Systems Inc., Facebook Inc., Microsoft Corp., Mimecast Ltd., MobileIron Inc., Oracle Corp., Rackspace Hosting Inc., Red Hat Inc., RingCentral, Salesforce.com Inc., VMware Inc., Workday Inc.. America-Telecom Equipment: ADTRAN Inc., ARRIS International Plc, Ciena Corp., Finisar Corp., Infinera Corp., Juniper Networks Inc., Lumentum Holdings. Asia Pacific Autos & Auto Parts: Amara Raja Batteries Ltd, Ashok Leyland, Bajaj Auto, Cheng Shin Rubber, Eicher Motors, Exide Industries, Hankook Tire, Hero MotoCorp, Hyundai Mobis, Hyundai Motor Co., Hyundai Wia, Kenda Rubber Industrial Co., Kia Motors, Mahindra & Mahindra, Maruti Suzuki India, Nexen Tire, Tata Motors, TVS Motor. Asia Pacific Chemicals: China Steel Chemical, Far Eastern New Century Corp., Formosa Chemicals & Fibre, Formosa Plastics, Hanwha Chemical, Kumho Petro Chemical Co., LG Chem, Lotte Chemical, Nan Ya Plastics, Petronas Chemicals Group, PTT Global Chemical, Taiwan Synthetic Rubber Corp..

Goldman Sachs Global Investment Research 80 April 13, 2016 Profiles in Innovation

Asia Pacific Energy: Bangchak Petroleum PCL, Bharat Petroleum, Cairn India Ltd., China Petroleum & Chemical (A), China Petroleum & Chemical (ADS), China Petroleum & Chemical (H), CNOOC, CNOOC (ADR), Formosa Petrochemical Corp., Gas Authority of India, Gujarat State Petronet, Hindustan Petroleum, Indian Oil Corp., Indraprastha Gas Ltd., IRPC PCL, OCI Co., Oil & Natural Gas Corp., Oil India, Perusahaan Gas, PetroChina (A), PetroChina (ADR), PetroChina (H), Petronet LNG, PTT Exploration and Production PCL, PTT Public Co., Reliance Industries, Reliance Industries (GDR), Thai Oil. Asia Pacific Infrastructure: China Communications Construction (A), China Communications Construction (H), China Machinery Engineering Corp., China Railway Construction (A), China Railway Construction (H), China Railway Group (A), China Railway Group (H), China State Construction Intl, Sinopec Engineering Group. Asia Pacific Technology: AAC Technologies, Acer, Advanced Semiconductor Engineering, ASUSTeK Computer, Casetek Holdings, Catcher Technology, Compal Electronics, Hon Hai Precision, Largan Precision, Lenovo Group, Pegatron, Quanta Computer, Siliconware Precision Industries, TPK Holding, Wistron. China Capital Goods: China Railway Signal & Communication, CRRC Corp. (A), CRRC Corp. (H), Estun Automation Co., Guangxi Liugong, Haitian International Holdings Limited, Han's Laser Technology Industry Group, Heli, Hollysys Automation Technologies Ltd., Huagong Tech Co., Lonking Holdings, Sany Heavy, Sany Heavy Equipment International, Shanghai STEP Electric Corporation, Shantui, Shenzhen Inovance Techology Co., Ltd, Siasun Robot&Automation Co., Ltd, Tian Di Science & Technology, Yangtze Optical Fibre and Cable, Zhengzhou Coal Mining Machinery, Zhuzhou CRRC Times Electric Co., Zoomlion (A), Zoomlion (H). Europe-Autos & Auto Parts: Autoliv Inc., BMW, CNH Industrial, CNH Industrial, Continental, Daimler AG, Faurecia, Fiat Chrysler Automobiles NV, Fiat Chrysler Automobiles NV, GKN, Hella KGaA Hueck, Michelin, Nokian Renkaat, Peugeot, Porsche, Renault, Valeo, Volkswagen, Volvo. Europe-IT Services: Atos, Capgemini, Indra, Wirecard, Worldline, Worldpay Group. Europe-Machinery & Elec Equip: ABB Ltd., Alfa Laval, Alstom, Assa Abloy B, Atlas Copco, Fincantieri SpA, FLSmidth & Co. A/S, Geberit Holding, KONE Corp., Legrand, Metso OYJ, Nexans, Outotec, Philips, Prysmian, Sandvik, Schindler Holding, Schneider Electric, Siemens AG, SKF, Wartsila, Weir Group. Europe-Software: Aveva, Dassault Systemes, Hexagon AB, Sage Group, SAP, SAP, Software AG, Temenos. Japan-Electronic Components: Alps Electric, Hirose Electric, Ibiden, IRISO Electronics, Japan Aviation Electronics Industry, Japan Display Inc., Kyocera, Mabuchi Motor, Minebea, Mitsumi Electric, Murata Mfg., NGK Insulators, NGK Spark Plug, Nichicon, Nidec, Nippon Ceramic, Nippon Chemi-Con, Nitto Denko, Pacific Industrial, Shinko Electric Industries, Taiyo Yuden, TDK. Japan-Machinery: Aida Engineering Ltd., Daikin Industries, Fanuc, Harmonic Drive Systems Inc., Hitachi Construction Machinery, JTEKT, Keyence, Komatsu, Kubota, Makita, NSK, NTN, Okuma Corp., SMC, THK, Tsubaki Nakashima Co., Yaskawa Electric. Taiwan Machinery: Advantech Corporation, Airtac International Group, Delta Electronics, Hiwin Corporation. Company-specific regulatory disclosures Compendium report: please see disclosures at http://www.gs.com/research/hedge.html. Disclosures applicable to the companies included in this compendium can be found in the latest relevant published research Distribution of ratings/investment banking relationships Goldman Sachs Investment Research global Equity coverage universe

Rating Distribution Investment Banking Relationships Buy Hold Sell Buy Hold Sell Global 32% 53% 15% 65% 58% 51% As of April 1, 2016, Goldman Sachs Global Investment Research had investment ratings on 3,029 equity securities. Goldman Sachs assigns stocks as Buys and Sells on various regional Investment Lists; stocks not so assigned are deemed Neutral. Such assignments equate to Buy, Hold and Sell for the purposes of the above disclosure required by the FINRA Rules. See 'Ratings, Coverage groups and views and related definitions' below. The Investment Banking Relationships chart reflects the percentage of subject companies within each rating category for whom Goldman Sachs has provided investment banking services within the previous twelve months. Price target and rating history chart(s) Compendium report: please see disclosures at http://www.gs.com/research/hedge.html. Disclosures applicable to the companies included in this compendium can be found in the latest relevant published research Regulatory disclosures Disclosures required by United States laws and regulations See company-specific regulatory disclosures above for any of the following disclosures required as to companies referred to in this report: manager or co-manager in a pending transaction; 1% or other ownership; compensation for certain services; types of client relationships; managed/co- managed public offerings in prior periods; directorships; for equity securities, market making and/or specialist role. Goldman Sachs trades or may trade as a principal in debt securities (or in related derivatives) of issuers discussed in this report. The following are additional required disclosures: Ownership and material conflicts of interest: Goldman Sachs policy prohibits its analysts, professionals reporting to analysts and members of their households from owning securities of any company in the analyst's area of coverage. Analyst compensation: Analysts are paid in part based on the profitability of Goldman Sachs, which includes investment banking revenues. Analyst as officer or director: Goldman Sachs policy prohibits its analysts, persons reporting to analysts or members of their households from serving as an officer, director, advisory board member or employee of any company in the analyst's area of coverage. Non-U.S. Analysts: Non-U.S. analysts may not be associated persons of Goldman, Sachs & Co. and therefore may not be subject to FINRA Rule 2241 or FINRA Rule 2242 restrictions on communications with subject company, public appearances and trading securities held by the analysts. Distribution of ratings: See the distribution of ratings disclosure above. Price chart: See the price chart, with changes of ratings and price targets in prior periods, above, or, if electronic format or if with respect to multiple companies which are the subject of this report, on the Goldman Sachs website at http://www.gs.com/research/hedge.html.

Goldman Sachs Global Investment Research 81 April 13, 2016 Profiles in Innovation

Additional disclosures required under the laws and regulations of jurisdictions other than the United States The following disclosures are those required by the jurisdiction indicated, except to the extent already made above pursuant to United States laws and regulations. Australia: Goldman Sachs Australia Pty Ltd and its affiliates are not authorised deposit-taking institutions (as that term is defined in the Banking Act 1959 (Cth)) in Australia and do not provide banking services, nor carry on a banking business, in Australia. This research, and any access to it, is intended only for "wholesale clients" within the meaning of the Australian Corporations Act, unless otherwise agreed by Goldman Sachs. In producing research reports, members of the Global Investment Research Division of Goldman Sachs Australia may attend site visits and other meetings hosted by the issuers the subject of its research reports. In some instances the costs of such site visits or meetings may be met in part or in whole by the issuers concerned if Goldman Sachs Australia considers it is appropriate and reasonable in the specific circumstances relating to the site visit or meeting. Brazil: Disclosure information in relation to CVM Instruction 483 is available at http://www.gs.com/worldwide/brazil/area/gir/index.html. Where applicable, the Brazil-registered analyst primarily responsible for the content of this research report, as defined in Article 16 of CVM Instruction 483, is the first author named at the beginning of this report, unless indicated otherwise at the end of the text. Canada: Goldman Sachs Canada Inc. is an affiliate of The Goldman Sachs Group Inc. and therefore is included in the company specific disclosures relating to Goldman Sachs (as defined above). Goldman Sachs Canada Inc. has approved of, and agreed to take responsibility for, this research report in Canada if and to the extent that Goldman Sachs Canada Inc. disseminates this research report to its clients. Hong Kong: Further information on the securities of covered companies referred to in this research may be obtained on request from Goldman Sachs (Asia) L.L.C. India: Further information on the subject company or companies referred to in this research may be obtained from Goldman Sachs (India) Securities Private Limited, Research Analyst - SEBI Registration Number INH000001493, 951-A, Rational House, Appasaheb Marathe Marg, Prabhadevi, Mumbai 400 025, India, Corporate Identity Number U74140MH2006FTC160634, Phone +91 22 6616 9000, Fax +91 22 6616 9001. Goldman Sachs may beneficially own 1% or more of the securities (as such term is defined in clause 2 (h) the Indian Securities Contracts (Regulation) Act, 1956) of the subject company or companies referred to in this research report. Japan: See below. Korea: Further information on the subject company or companies referred to in this research may be obtained from Goldman Sachs (Asia) L.L.C., Seoul Branch. New Zealand: Goldman Sachs New Zealand Limited and its affiliates are neither "registered banks" nor "deposit takers" (as defined in the Reserve Bank of New Zealand Act 1989) in New Zealand. This research, and any access to it, is intended for "wholesale clients" (as defined in the Financial Advisers Act 2008) unless otherwise agreed by Goldman Sachs. Russia: Research reports distributed in the Russian Federation are not advertising as defined in the Russian legislation, but are information and analysis not having product promotion as their main purpose and do not provide appraisal within the meaning of the Russian legislation on appraisal activity. Singapore: Further information on the covered companies referred to in this research may be obtained from Goldman Sachs (Singapore) Pte. (Company Number: 198602165W). Taiwan: This material is for reference only and must not be reprinted without permission. Investors should carefully consider their own investment risk. Investment results are the responsibility of the individual investor. United Kingdom: Persons who would be categorized as retail clients in the United Kingdom, as such term is defined in the rules of the Financial Conduct Authority, should read this research in conjunction with prior Goldman Sachs research on the covered companies referred to herein and should refer to the risk warnings that have been sent to them by Goldman Sachs International. A copy of these risks warnings, and a glossary of certain financial terms used in this report, are available from Goldman Sachs International on request. European Union: Disclosure information in relation to Article 4 (1) (d) and Article 6 (2) of the European Commission Directive 2003/125/EC is available at http://www.gs.com/disclosures/europeanpolicy.html which states the European Policy for Managing Conflicts of Interest in Connection with Investment Research. Japan: Goldman Sachs Japan Co., Ltd. is a Financial Instrument Dealer registered with the Kanto Financial Bureau under registration number Kinsho 69, and a member of Japan Securities Dealers Association, Financial Futures Association of Japan and Type II Financial Instruments Firms Association. Sales and purchase of equities are subject to commission pre-determined with clients plus consumption tax. See company-specific disclosures as to any applicable disclosures required by Japanese stock exchanges, the Japanese Securities Dealers Association or the Japanese Securities Finance Company. Ratings, coverage groups and views and related definitions Buy (B), Neutral (N), Sell (S) -Analysts recommend stocks as Buys or Sells for inclusion on various regional Investment Lists. Being assigned a Buy or Sell on an Investment List is determined by a stock's return potential relative to its coverage group as described below. Any stock not assigned as a Buy or a Sell on an Investment List is deemed Neutral. Each regional Investment Review Committee manages various regional Investment Lists to a global guideline of 25%-35% of stocks as Buy and 10%-15% of stocks as Sell; however, the distribution of Buys and Sells in any particular coverage group may vary as determined by the regional Investment Review Committee. Regional Conviction Buy and Sell lists represent investment recommendations focused on either the size of the potential return or the likelihood of the realization of the return. Return potential represents the price differential between the current share price and the price target expected during the time horizon associated with the price target. Price targets are required for all covered stocks. The return potential, price target and associated time horizon are stated in each report adding or reiterating an Investment List membership. Coverage groups and views: A list of all stocks in each coverage group is available by primary analyst, stock and coverage group at http://www.gs.com/research/hedge.html. The analyst assigns one of the following coverage views which represents the analyst's investment outlook on the coverage group relative to the group's historical fundamentals and/or valuation. Attractive (A). The investment outlook over the following 12 months is favorable relative to the coverage group's historical fundamentals and/or valuation. Neutral (N). The investment outlook over the following 12 months is neutral relative to the coverage group's historical fundamentals and/or valuation. Cautious (C). The investment outlook over the following 12 months is unfavorable relative to the coverage group's historical fundamentals and/or valuation. Not Rated (NR). The investment rating and target price have been removed pursuant to Goldman Sachs policy when Goldman Sachs is acting in an advisory capacity in a merger or strategic transaction involving this company and in certain other circumstances. Rating Suspended (RS). Goldman Sachs Research has suspended the investment rating and price target for this stock, because there is not a sufficient fundamental basis for determining, or there are legal, regulatory or policy constraints around publishing, an investment rating or target. The previous investment rating and price target, if any, are no longer in effect for this stock and should not be relied upon. Coverage Suspended (CS). Goldman Sachs has suspended coverage of this company. Not Covered (NC). Goldman Sachs does not cover this company. Not Available or Not Applicable (NA). The information is not available for display or is not applicable. Not Meaningful (NM). The information is not meaningful and is therefore excluded. Global product; distributing entities The Global Investment Research Division of Goldman Sachs produces and distributes research products for clients of Goldman Sachs on a global basis. Analysts based in Goldman Sachs offices around the world produce equity research on industries and companies, and research on macroeconomics, currencies, commodities and portfolio strategy. This research is disseminated in Australia by Goldman Sachs Australia Pty Ltd (ABN 21 006 797 897); in Brazil by Goldman Sachs do Brasil Corretora de Títulos e Valores Mobiliários S.A.; in Canada by either Goldman Sachs Canada Inc. or Goldman, Sachs & Co.; in Hong Kong by Goldman Sachs (Asia) L.L.C.; in India by Goldman Sachs (India) Securities Private Ltd.; in Japan by Goldman Sachs Japan Co., Ltd.; in the Republic of Korea by Goldman Sachs (Asia) L.L.C., Seoul Branch; in New Zealand by Goldman Sachs New Zealand Limited; in Russia by OOO Goldman Sachs; in Singapore by Goldman Sachs (Singapore) Pte. (Company Number: 198602165W); and in

Goldman Sachs Global Investment Research 82 April 13, 2016 Profiles in Innovation

the United States of America by Goldman, Sachs & Co. Goldman Sachs International has approved this research in connection with its distribution in the United Kingdom and European Union. European Union: Goldman Sachs International authorised by the Prudential Regulation Authority and regulated by the Financial Conduct Authority and the Prudential Regulation Authority, has approved this research in connection with its distribution in the European Union and United Kingdom; Goldman Sachs AG and Goldman Sachs International Zweigniederlassung Frankfurt, regulated by the Bundesanstalt für Finanzdienstleistungsaufsicht, may also distribute research in Germany. General disclosures This research is for our clients only. Other than disclosures relating to Goldman Sachs, this research is based on current public information that we consider reliable, but we do not represent it is accurate or complete, and it should not be relied on as such. The information, opinions, estimates and forecasts contained herein are as of the date hereof and are subject to change without prior notification. We seek to update our research as appropriate, but various regulations may prevent us from doing so. Other than certain industry reports published on a periodic basis, the large majority of reports are published at irregular intervals as appropriate in the analyst's judgment. Goldman Sachs conducts a global full-service, integrated investment banking, investment management, and brokerage business. We have investment banking and other business relationships with a substantial percentage of the companies covered by our Global Investment Research Division. Goldman, Sachs & Co., the United States broker dealer, is a member of SIPC (http://www.sipc.org). Our salespeople, traders, and other professionals may provide oral or written market commentary or trading strategies to our clients and principal trading desks that reflect opinions that are contrary to the opinions expressed in this research. Our asset management area, principal trading desks and investing businesses may make investment decisions that are inconsistent with the recommendations or views expressed in this research. The analysts named in this report may have from time to time discussed with our clients, including Goldman Sachs salespersons and traders, or may discuss in this report, trading strategies that reference catalysts or events that may have a near-term impact on the market price of the equity securities discussed in this report, which impact may be directionally counter to the analyst's published price target expectations for such stocks. Any such trading strategies are distinct from and do not affect the analyst's fundamental equity rating for such stocks, which rating reflects a stock's return potential relative to its coverage group as described herein. We and our affiliates, officers, directors, and employees, excluding equity and credit analysts, will from time to time have long or short positions in, act as principal in, and buy or sell, the securities or derivatives, if any, referred to in this research. The views attributed to third party presenters at Goldman Sachs arranged conferences, including individuals from other parts of Goldman Sachs, do not necessarily reflect those of Global Investment Research and are not an official view of Goldman Sachs. Any third party referenced herein, including any salespeople, traders and other professionals or members of their household, may have positions in the products mentioned that are inconsistent with the views expressed by analysts named in this report. This research is not an offer to sell or the solicitation of an offer to buy any security in any jurisdiction where such an offer or solicitation would be illegal. It does not constitute a personal recommendation or take into account the particular investment objectives, financial situations, or needs of individual clients. Clients should consider whether any advice or recommendation in this research is suitable for their particular circumstances and, if appropriate, seek professional advice, including tax advice. The price and value of investments referred to in this research and the income from them may fluctuate. Past performance is not a guide to future performance, future returns are not guaranteed, and a loss of original capital may occur. Fluctuations in exchange rates could have adverse effects on the value or price of, or income derived from, certain investments. Certain transactions, including those involving futures, options, and other derivatives, give rise to substantial risk and are not suitable for all investors. Investors should review current options disclosure documents which are available from Goldman Sachs sales representatives or at http://www.theocc.com/about/publications/character-risks.jsp. Transaction costs may be significant in option strategies calling for multiple purchase and sales of options such as spreads. Supporting documentation will be supplied upon request. All research reports are disseminated and available to all clients simultaneously through electronic publication to our internal client websites. Not all research content is redistributed to our clients or available to third-party aggregators, nor is Goldman Sachs responsible for the redistribution of our research by third party aggregators. For research, models or other data available on a particular security, please contact your sales representative or go to http://360.gs.com. Disclosure information is also available at http://www.gs.com/research/hedge.html or from Research Compliance, 200 West Street, New York, NY 10282. © 2016 Goldman Sachs. No part of this material may be (i) copied, photocopied or duplicated in any form by any means or (ii) redistributed without the prior written consent of The Goldman Sachs Group, Inc.

Goldman Sachs Global Investment Research 83