Choose freedom. Why solving the prisoner s dilemma is the key to a successful future for IP deal making

Issue 84 July/August 2017 Choose freedom Why solving the prisoner s dilemma is the key to a successful future for IP deal making Latest benchmarking survey reveals optimism in the US Xiaomi s plan to put patents at the heart of its strategy Calculating your RoI when joining a defensive aggregator The inside story on the great InterDigital comeback How plaintiffs can develop winning Chinese campaigns

40 Feature Inside the IoT IoT markets and their patent landscapes The world of the Internet of Things is becoming increasingly complex in terms of both markets and technology definitions. And that is before we even get to the patents By Ian MacLean, Raymond Angers and Goran Grbic Trying to understand and anticipate how the Internet of Things (IoT) IP landscape will evolve is still an imprecise art. Between 2014 and 2016, Gartner revised its IoT market projections significantly. While some numbers rose and others fell, it seems clear that the overall market is still a large one. From a patent ownership and development standpoint, other trends are starting to emerge which could have a significant impact on what the IoT landscape will look like further out. This article examines some of the patent activity in the different IoT segments, taking a broad perspective. We observe that growth for patents in the areas of things and analytics is up, with many of the leaders being as expected; from an M&A perspective, this is still an active market. We also observe that this activity is more about technology than intellectual property. To begin with, let us set some parameters around defining those segments and look at the growth projections for particular markets. The four main technology segments in IoT and their definitions are illustrated in Figure 1. However, the lines defining the types of product which belong to each segment are not always clear, particularly for the things segment. For example, a smartphone can be a thing (eg, an always-on microphone), but it can also be a gateway which communicates with a smartwatch via Bluetooth and relays the data back to the network. A thing may be a standalone device, such as a heart rate monitor chest strap or a smart lightbulb, or a complex device such as an automobile, with many sensors and cellular connectivity. In addition to data being sent from things to the network, things can communicate directly with each other in a peer-to-peer fashion or by using a gateway as an intermediary. The types of product that fall under each category will depend on the ecosystem of a particular application. IoT market growth We compared Gartner s 2016 projections to those from 2014 (Figures 2 to 5) and focused on the smart home and automotive segments. In 2014 the largest shipments were forecast for the smart home and automotive segments. The highest growth rate then was within the medical devices segment, followed by automotive. This changed somewhat in 2016. The growth expectations for medical were radically reduced, while those for the automotive segment are forecast to increase over the next five years, fuelled by self-driving and connected cars. The industrial segment was also projected as having more growth in the 2016 projections than in 2014. The overall IoT unit shipment forecast through has been reduced by roughly 17%, compared to 2014 projections. This is due predominantly to reduced smart home projections, although this segment still has the highest level of IoT shipments because of the sheer number of electronic products which can be included as part of a comprehensive connected home. Additionally, the forecast for industrial over the next few years remains positive. The 2014 and 2016 market shipment forecasts are extremely similar starting in 2016 and going through. In both forecasts, the overall unit shipment charts show a 17% decrease by. Both the smart home and industrial segments are complex in composition and require a deeper look. The smart home market projections are clearer when the market is divided into two segments: entertainment and home automation and management (Figures 6 and 7). The home automation and management segment will drive substantial growth in the smart home market. This covers everything in the home with sensors and network modules used to provide anywhere, anytime controlling and monitoring ability, and may include appliances, surveillance, lighting and temperature control systems. While a large variety of products are available, consumers and industry alike are still trying to determine where the actual need resides. This part of the smart home market is in a rapid growth stage. It looks set to eventually surpass the entertainment segment shipments and dominate the entire smart home space. Let us look at the manufacturing segment (Figures 8 and 9). While an increasing number of sensors will be introduced in the industrial market, development will likely be incremental, with well-understood use cases; hence, the growth rate in this segment is much lower than in other IoT market segments. The evolution in industrial may come from wearables which are designed to provide field workers with realtime communication for remote diagnostic support. For example, miners wearing smart glasses with video transmission could allow above-ground workers to see what they see underground, thereby shortening fieldproblem resolution times. However, such applications are at a nascent state.

Inside the IoT Feature 41 FIGURE 1. Main technology segments Things Things sense environmental conditions and produce data which is sent to network platforms and servers for analytic generation Gateway/network Gateways communicate with multiple things and the network provides the transport of data to the cloud Compute/storage Compute/storage is the infrastructure where the data is received, stored and processed Analytics Analytics is the software layer that analyses the data, presents it to the user, and interacts with other layers based on analysis and user input FIGURE 2. Year-on-year growth rate forecast by shipment 2014 FIGURE 3. Year-on-year growth rate forecast by shipment 2016 YoY rate of growth 120% 100% 80% 60% 40% Automotives Health and fitness YoY rate of growth 120% 100% 80% 60% 40% Automotives Health and fitness 20% 20% 0% 0% FIGURE 4. Market shipment forecast 2014 FIGURE 5. Market shipment forecast 2016 Millions of units 6,000 5,000 4,000 3,000 2,000 Health and fitness Automotives Millions of units 6,000 5,000 4,000 3,000 2,000 Health and fitness Automotives 1,000 1,000 0 0 FIGURE 6. Year-on-year growth rate forecast by shipment 2014 versus 2016 FIGURE 7. sub-markets shipment forecast 2014 versus 2016 YoY rate of growth 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2016 entertainment 2016 home automation and mgmt 2014 entertainment 2014 home automation and mgmt Millions of units 3,000 2,500 2,000 1,500 1,000 500 0 2014 entertainment 2016 entertainment 2014 home automation and mgmt 2016 home automation and mgmt

42 Feature Inside the IoT Patenting activity in different market spaces Early in 2015, TechInsights analysed 38 companies active in IoT and their patenting activity in four specific IoT segments. For this article, we updated the analysis on the original set and for 2017 added a similar list of companies, some of which have since been acquired. Table 1 lists the number of patents for each of the companies in each of the four IoT technology segments. Using publicly available tools, we retrieved the US portfolios (granted patents and pending applications) with title, abstract and independent claims for each company. Next, using a proprietary tool, we compared these portfolios to classifiers created by subject-matter experts, placing each patent in the best match among the four IoT segments. There is some subjectivity in the specific placement of each patent in one segment or another, considering that there is a level of accuracy and recall in any analytic investigation as well as the blurred boundaries already mentioned between some segments. In 2015 Samsung, IBM, Microsoft, Qualcomm, Nokia and Intel ranked as the top six companies in terms of number of patents in the IoT space. Also in 2015 a strong percentage of Samsung s, Qualcomm s and Intel s IoT patent portfolios lay in the gateway/ network spaces; while IBM and Microsoft IoT patents were concentrated on the compute/storage spaces, since technology in these segments was established well before IoT became so prevalent. Many companies with large patent portfolios which were also big players in the traditional mobile and data centre markets are now expanding into IoT. The company with the highest number of things patents was Samsung, which had double those of the crowded pack following it including Panasonic, Microsoft and Qualcomm. The entity with the highest number of analytics patents was Microsoft, followed closely by IBM and Alphabet (Google), then Sony, Samsung and Facebook. Looking at 2017, we see that from the top 20 from our 2015 list, analytics had the largest growth at 17%, followed closely by gateway/network at 16%, with things at 13% and compute/storage at 3%. The group with the largest overall number of patents among the 20 is compute/storage, followed by gateway/network. The top six companies continue to be well represented across all four segments. Samsung, IBM and Microsoft remain at the top, while Intel moved up to the number four spot, with healthy increases in things, gateway/network and compute/storage. LG Electronics and Google both leapfrogged over Nokia the former based on strong growth in gateway/network and the latter with consistent innovation efforts in all four areas. Of our top 10 in 2017, only Sony s and Nokia s overall portfolios dropped from 2015 according to our IoT model. The leaders in each segment have not changed: Samsung in things, Qualcomm in gateway/network, IBM in compute/ storage and Microsoft in analytics. Growth over all four segments was consistent for Samsung (12%), IBM (11%) and Microsoft (11%), with more significant growth being observed for Intel (22%) and LG (24%). Cisco s acquisition of Jasper an industry-leading cloud-based IoT service platform provider means that it is now well positioned to look towards the compute/ storage space. FIGURE 8. Year-on-year growth rate forecast by shipment 2014 versus 2016 YoY rate of growth Millions of units 30% 25% 20% 15% 10% 5% 0% FIGURE 9. sub-markets shipment forecast 2014 versus 2016 250 200 150 100 50 0 2014 manufacturing and natural resources 2016 manufacturing and natural resources 2014 utilities 2016 utilities Amazon has a significant number of patents, specifically in the compute/storage and analytics segments. It scored a significant success in the things area with its Echo, which has been referenced as a step forward in consumer acceptance of IoT devices. It will be worth watching to see whether that growth continues. With its purchase of ARM, Softbank is a strong newcomer. However, from a patent perspective, more intellectual property came from Softbank s legacy rather than from the existing ARM portfolio. Its acquisition of ARM does mean that it has acquired technology which bolsters its position with some other large traditional IoT players, especially in the areas of low-power mobile system-on-a-chip cores. TDK s purchases of Micronas in 2015 and InvenSense in 2017 are obvious steps towards it becoming a more important player in IoT. By adding InvenSense, TDK has strengthened its position in new sensor solutions in IoT, automotive and information and communications technology; it also brings a significant number of IoT assets, particularly in the things space. All of this makes it a more prominent player in IoT intellectual property. An interesting and still developing situation is that 2016 manufacturing and natural resources 2016 utilities 2014 manufacturing and natural resources 2014 utilities

Inside the IoT Feature 43 of Thoma Bravo, LLC, a US private equity and growth capital firm which has raised more than $1.5 billion in investor commitments since 2008. Compuware, which it acquired in 2014, was a leader in both application performance monitoring and mainframe solutions. Riverbed provided the application performance infrastructure. Both were involved in the compute/ storage segments. Compuware and Riverbed added guided analytics, while the purchase of Qlik added visual analytics, delivering intuitive solutions for selfservice data visualisation. As a result, Thoma Bravo has positioned itself as a presence in IoT intellectual property, especially in compute/storage and analytics. It will be interesting to see what its next move will be. As was recently announced, Renesas has used the Cadence Perspec System Verifier to verify its new microcontroller unit design for IoT applications. Renesas microcontrollers have more IP blocks and TABLE 1. Comparison of 2015 and 2017 company ranking of IoT patents in four IoT segments 2015 2017 Companies Things Gateway/ network Compute/ storage Analytics Grand total Companies Things Gateway/ network Compute/ storage Analytics Grand total Samsung 3351 6203 4384 1204 15142 Samsung 3865 7042 4852 1209 16968 IBM 1452 466 10296 2740 14954 IBM 1582 603 11000 3370 16555 Microsoft 1739 574 8375 3221 13909 Microsoft 1996 783 8849 3772 15400 Qualcomm 1573 7006 1479 572 10630 Intel 1868 4202 5076 714 11860 Nokia / ALU 1162 5032 2835 726 9755 Qualcomm 1805 7884 1516 564 11769 Intel 1466 3260 4402 614 9742 LG Electronics 895 6887 1413 684 9879 Alphabet 1347 2128 2888 2676 9039 Alphabet (Google) 1604 2200 3090 2983 9877 LG 763 5270 1307 643 7983 Nokia 1163 5301 2443 618 9525 Sony 1667 1775 2361 1384 7187 Ericsson 615 5734 1427 220 7996 Ericsson 533 4888 1439 246 7106 Sony 1711 2010 2109 1263 7093 Panasonic 1964 1933 1641 773 6311 Apple 1914 2107 1832 696 6549 Cisco 1090 934 3200 664 5888 Cisco 1230 1239 3265 637 6371 Apple 1489 1635 1718 648 5490 Panasonic 1846 1614 1474 735 5669 GE 1300 441 1473 893 4107 GE 1261 509 1308 904 3982 Bosch 1439 159 1105 491 3194 Bosch 1500 177 1060 528 3265 TI 984 982 703 175 2844 Amazon 437 171 1729 908 3245 Philips 940 560 405 470 2375 TI 1034 1003 687 153 2877 NXP / Freescale 626 559 704 45 1934 Softbank/ARM 269 1543 779 172 2763 ST Micro 791 339 721 64 1915 Philips 1074 663 415 496 2648 Amazon 191 89 1006 608 1894 Facebook 97 91 643 1573 2404 Facebook 57 35 554 1095 1741 NXP / Freescale 736 752 786 43 2317 Renesas 367 270 667 47 1351 STMicro 894 352 718 67 2031 Mediatek 180 654 188 27 1049 Mediatek 205 838 190 25 1258 Infineon 326 209 339 35 909 Renesas 344 247 607 42 1240 Salesforce 20 4 538 184 746 Infineon 348 247 320 35 950 Gemalto 150 49 254 15 468 Microchip/Atmel 230 220 390 11 851 Microchip 120 81 243 7 451 Salesforce 17 10 477 191 695 Atmel 99 133 142 5 379 Gemalto 189 96 297 16 598 Alibaba 5 12 122 110 249 Arris/Pace 67 102 218 79 466 ARM 53 10 121 2 186 TDK/Micronas 167 106 88 64 425 Pioneer 36 21 32 43 132 Alibaba 15 17 225 138 395 Invensense 107 2 5 1 115 ThomaBravo/Qlik 31 0 139 42 212 Pace 13 16 15 6 50 Invensense 166 3 4 4 177 Splunk 5 0 16 24 45 Splunk 12 0 54 85 151 Baidu 0 0 11 4 15 Pioneer 30 16 21 28 95 Micronas 10 0 2 0 12 Baidu 4 0 33 13 50 Tableau 0 0 0 11 11 Tableau 0 0 0 19 19 Qlik 0 0 0 6 6

44 Feature Inside the IoT FIGURE 10. TDK s acquisition of InvenSense TDK InvenSense FIGURE 11. Qualcomm s takeover bid for NXP Qualcomm NXP InvenSense added sensors NXP enhanced three IoT segments complex sub-systems than existing designs, resulting in thousands of combinations of potential access conflicts. The previous Renesas approach involved a manual process for creating use-case test scenarios, which was time consuming due to the large number of combinations which needed to be verified. By replacing its legacy process with the Perspec System Verifier, Renesas achieved the benefits of an efficient, algorithmbased system level verification solution which enabled the automatic generation of complex test scenarios. From an IP perspective, its influx of new assets is holding steady. It is expected that more innovation will come from the microcontroller testing environment, with less emphasis on gateway/network activity. M&A activity Now let us turn our attention to some of the acquisition activity which has taken place in each of the four segments and observe the portfolio impact of those mergers in more detail. Each of the patent landscape plots was created using patents from each company. Each dot (Figure 10 to 13) represents a patent, while the three-word clusters around the peaks represent the most common words from patents in those peaks. Things TDK s acquisition of InvenSense, announced in December 2016, is an example of a company focused on things strengthening its presence in the technology segment (Figure 10). This can be seen as a continuation of TDK s strategy, building on the markets for Micronas magnetic sensors. InvenSense is focused on microelectrical mechanical systems sensors such as gyroscopes, accelerometers and microphones which are increasingly found in a variety of IoT applications (eg, wearables, audio control). Acquisitions such as these are driving TDK s transition from a smartphone component supplier to an IoT leader. The patent landscape shows a significant cluster of intellectual property related to microphones and sensors which the InvenSense acquisition brings to the TDK IP portfolio (approximately 500 InvenSense US publications). Gateway/network Qualcomm, the leader in mobile communication processors, has the highest proportion of its IoTrelated intellectual property in the gateway/network technology segments. While its massive $47 billion takeover bid of NXP, announced last October, is not expected to close until the end of 2017, the news came hot on the heels of NXP s acquisition of Freescale Semiconductor in late 2015. The combined NXP/ Freescale is now a leader in both the automotive and near-field communication markets. The intellectual property of the new entity spans the things, gateway/ network and compute/storage IoT segments, with all three now appearing to be improved by the NXP portfolio (Figure 11). Compute/storage Cisco, the world s largest networking company, has long held a strong presence in the compute/storage and gateway/network IoT technology segments. The $1.4 billion acquisition of Jasper in March 2016 is an example of an equipment supplier expanding into the IoT services markets (Figure 12). Jasper s control centre platform is a cloud-based solution which can be configured to meet specialised enterprise needs across a wide range of business models, technologies and industries. For example, Jasper s connected car cloud is used by 11 automotive original equipment manufacturers worldwide, including General Motors, Nissan, Ford Motor Company, Volkswagen and Tesla Motors. There are many examples of established IoT players such as Cisco acquiring small focused technology companies such as Jasper with modest patent holdings. While the patent landscape suggests that Cisco has significant intellectual property in the client-server

Inside the IoT Feature 45 FIGURE 12. Cisco s acquisition of Jasper Cisco Systems Jasper FIGURE 13. Softbank s takeover of ARM ARM Softbank Jasper s IP was focused ARM complemented Softbank area at a high level, Jasper s unique implementation and market adoption have made it a valuable addition to Cisco s IoT product offering. Analytics Japanese telecommunications and internet corporation Softbank completed its $31 billion takeover of UKbased chip designer ARM Holdings last autumn (Figure 13). The main motivation for this according to Softbank CEO Masayoshi Son was to capture the very significant opportunities provided by the Internet of Things. Looking at the IP analytics, we found this to be more of a technical play than IP play. Our analysis of the deal presented in an earlier blog posted on October 2016 indicated that of ARM s 2,300 US and UK patents, fewer than 10% were specific to IoT. We highlight IoT-related assets only in Figure 13 to illustrate this point. While the Softbank acquisition of ARM significantly increased the portfolio size, only a small proportion of ARM patents apply directly to IoT. Ones to watch Gemalto in things As a company basing its main products on security, Gemalto s IoT solution comes from various sources. First is the Cinterion Secure Element, a tamper- While the patent landscape suggests that Cisco has significant intellectual property in the client-server area at a high level, Jasper s unique implementation and market adoption have made it a valuable addition to Cisco s IoT product offering resistant embedded hardware component which allows devices to securely connect to the network an integral part of IoT things solutions. Gemalto also has important elements relating to compute/storage, including its enterprise data encryption and cloud security, along with identity management, software monetisation and encryption key management. These areas where Gemalto has developed historically combined with its jump in things assets make it an attractive prospect. Alibaba in compute/storage and analytics Alibaba is partnering with Gemalto to help secure China s IoT ecosystem. Gemalto will provide connectivity and security for YunOS, a cloud-based IoT operating system developed by Alibaba. This alliance sits well with Alibaba s IP assets in IoT compute/storage and analytics. Splunk in compute/storage Splunk has announced its integration with ForeScout, whose products help companies to detect IoT devices on their networks. ForeScout entry does not bolster Splunk s IoT portfolio significantly, but does solidify the security portion of its IoT endpoint and network business. It appears that Splunk had bolstered its intellectual property in IoT compute/storage, which is not its traditional business space. Arris/Pace in compute/storage Arris/Pace is much stronger in all areas since merging. The driver behind the move was to put Arris in a stronger position relative to Cisco in the area of video infrastructure. The combined Arris/Pace now dominates the client hardware segments, while still lagging behind Cisco in network hardware and video infrastructure software. Although Pace did not bring a significant number of IP IoT assets to the deal, the combination has resulted in an interesting IoT portfolio in compute/storage.

46 Feature Inside the IoT FIGURE 14. 2016 IoT market position on industry lifecycle Sales ($) Health & fitness Automotive Time Pioneer Rapid growth Maturity growth Stabilisation Decline Tableau in analytics Given the overall growth in analytics, Tableau presents an interesting opportunity. Although still relatively inactive in pursuing patents, it could be attractive as a technology addition for a company interested in data visualisation and merging information from multiple data sources. Lifecycle of different IoT markets Figure 14 shows the current position of each IoT market in its own industry lifecycle. As the duration of each stage and the peak sales level differ for each market, the figure above does not reflect any absolute or relative sales, revenue level or timeline for IoT markets. The automotive segment has exited the pioneer stage and is now fully in the rapid growth stage. In the automotive segment, embedded sensors and communication modules are leading the growth; while the connected car segment is still early in its development, but with high expectations. are still at the pioneer phase, with substantially high growth rate but low volume. This could be related to the typically longer industry lifecycle in the medical segment. The smart home segment position has advanced a little towards the maturity growth stage, but it is still experiencing rapid growth. The home automation and management sub-segment is still the exception, with more than 70% year-on-year growth. Platform consolidation will help the growth of the smart home segment. The health and fitness segment is now fully in the maturity growth rate and marching quickly towards stabilisation. This is reflected in the proliferation of smart fitness products currently on the market. The industrial segment is the most mature of all IoT segments and is looking as though it will remain so for some time to come. What can we learn? Market growth and unit shipment projections are a moving target. This makes it tricky to understand exactly when mass acceptance will be realised and what the journey to get there looks like. Although the forecast for overall unit shipments may have dropped, IoT is still a significant market with keen interest from many big Action plan In the increasingly interconnected world of the Internet of Things (IoT), rights holders can take a few strategic steps to ensure that they are best positioned for market s ongoing growth and development: Understand your company s position in the IoT and be aware of which of the four segments your patents are best suited to. Monitor growth in the key markets, smart home, medical and automotive segments. Monitor the activities of those in adjacent markets or technologies which align with your strategic vision and could have a positive impact on where you wish to be positioned. Map your competition s patents to better understand their position. Map their patents and patent applications by technology groups and over time understand where they are focusing their innovation efforts. semiconductor and electronics businesses. Companies are actively patenting and looking to strengthen their market position through acquisitions. One area creating significant interest is platforms, which are described as overcrowded by IoT analytics report IoT 2016 in review: the eight most relevant IoT developments of the year, which states that at last count there were over 400 different platforms. Platform issues affect speed of adoption and interoperability, and standards in this area have yet to stabilise. Other factors that will affect the overall IoT landscape and that we have not touched on here include security implications and the legal hurdles around divided infringement, claim scope and subject-matter eligibility. It all adds up to continuing interest from companies which are seeking to ensure they are well positioned from a patent and underlying technology standpoint to leverage this dynamic, as yet not fully realised market. Ian MacLean is vice president, IP services, Raymond Angers is manager, IP technology solutions and Goran Grbic is IP solutions analyst at TechInsights, Ottawa, Canada