Economics & Statistics Series
Working Paper No. 30
Breakthrough technologies –
Robotics, innovation and intellectual property
C. Andrew Keisner
Breakthrough Technologies – Robotics, Innovation and Intellectual
C. Andrew Keisner∗, Julio Raffo∗∗, Sacha Wunsch-Vincent∗∗
Robotics technology and the increasing sophistication of artificial intelligence are
breakthrough innovations with significant growth prospects and the potential to disrupt
existing economic and social facets of everyday life. Few studies have analyzed the
developments of robotics innovation. This paper closes this gap by analyzing how
innovation in robotics is taking place, how it diffuses, and what role intellectual property (IP)
plays. The paper finds that robotics clusters are mainly located in the US, Europe, but
increasingly also in the Republic of Korea and China. The robotics innovation ecosystem
builds on cooperative networks of actors, including individuals, research institutions, and
firms. Governments play a significant role in supporting robotics innovation, in particular
through funding, military demand, and national robotics strategies. Robotics competitions
and prizes provide for an important incentive to innovation. Patents are used to exclude
third parties, to secure freedom to operate, to license technologies and to avoid litigation.
The countries with the highest number of filings are Japan, China, Republic of Korea and the
US. The growing stock of patents owned by universities and PROs, in particular in China, is
noteworthy too. Automotive and electronics companies are still the largest patent filers, but
new actors in fields such as medical technologies and the Internet are emerging. Secrecy is
often used as a tool to appropriate innovation. Copyright protection is relevant to robotics
too, mainly in its role in protecting software, and more recently in protecting so-called
Netlists. Finally, proprietary approaches co-exist with open-source robotics platforms which
are developing rapidly in robotics clusters.
Keywords: robotics, robot, artificial intelligence, innovation, patent, trade secret, copyright
JEL Classification: F23, L86, O3, L6
The views expressed in this article are those of the authors and do not necessarily reflect the
views of the World Intellectual Property Organization or its member states.
∗ At the time of drafting the study commissioned by the World Intellectual Property Organization: Attorney, Davis
& Gilbert LLP, New York, New York, U.S.A.
∗∗ Economics and Statistics Division, World Intellectual Property Organization, Geneva.
Prof. Roland Y. Siegwart, Swiss Federal Institute of Technology, ETH Zürich deserves
thanks for reading drafts this study and giving advice as to the future of the robotics industry.
A first draft of this study has been presented at a WIPO Workshop for the preparation of the
”World Intellectual Property Report 2015” in Geneva – February 5 and 6, 2015 and received
valuable feedback from workshop participants, in particular from Roger Burt, Remy Glaisner,
Deven Desai, Thomas Hören, and David Mowery. Comments were also received by Mirko
Boehm. Data were made available by the International Federation for Robotics, and by
Frank Tobe, The Robot Report.
Richard Corken, Christopher Harrison, and Marian Lilington of the UK IP Office provided
advice for the creation of the patent landscapes. Carsten Fink, Chief Economist, WIPO and
Francesca Guadagno, affiliated with WIPO during the report drafting, provided critical input
throughout the project.
Parts of this working paper are the result of a study that was commissioned by WIPO from
C. Andrew Keisner as background for the World Intellectual Property Report 2015:
Breakthrough Innovation and Economic Growth, as in WIPO (2015).
At bottom, robotics is about us. It is the discipline of emulating our lives, of wondering how
Director of the Laboratory for Perceptual Robotics
University of Massachusetts Amherst
The looming sophistication of robots and artificial intelligence (AI) and its consequences are
currently the subject of numerous debates. The fact that humanoid robots have recently
been trialed in supermarkets, schools, hospitals and retirement homes in the Europe, the
United States and Japan has given the field of robotics new prominence.
Technologists, economists, lawyers and other disciplines are speculating about the potential
uses, the social and the economic impacts of robotics innovation. In economic circles, the
debate often focusses on the - the potential positive or negative - employment impacts of
robots. Social scientists are debating the social influence of artificial companions.
Hollywood movies such as Ex_Machina or Her, too, have put the spotlight on the potential
upcoming superiority of AI which might increasingly rival human intelligence. All observers
agree that the pervasive uptake and impacts of robotics innovation is imminent, and
Yet, despite the central attention devoted to this expanding field of technology, few studies
have analyzed the developments of robotics innovation, and the underlying innovation eco-
system. Moreover, while the role of intellectual property (IP) is analyzed with respect to
numerous high-technology fields, such as information-, nano-, or bio-technology, literally no
studies are devoted to the use and uptake of various forms of IP for robotics innovation. The
few articles devoted to robotics innovation in prominent innovation journals date back to the
This paper aims to fill this gap by providing an up-to-date analysis of the robotics innovation
system, and the corresponding role of IP. Section 1 describes the history of robotics
innovation. Section 2 assesses its underlying economic contribution. Section 3 describes
the robotics innovation ecosystem. Section 4 analyzes the uptake and relevance of different
strands of IP to robotics innovation. Original patent landscapes for robotics are presented,
which also highlight the past and current epicenters of robotics innovation. A technical
annex sets out the computational approach taken to elaborate these patent statistics and
This is part of a broader series of studies for WIPO’s World IP Report 2015 as in WIPO
(2015) exploring the concrete linkages between innovation, IP, and growth in six areas of
breakthrough innovation (airplanes, antibiotics, semiconductors, 3D printing, nano-
technology and robotics).
1 See for example in Research Policy, see Kumaresan and Miyazaki (1999).
1. The Development of Robotics and their Economic Importance
Robotics is the field of technology which drives the development of robots for application in
areas as diverse as car factories, construction sites, schools, hospitals and private homes.
Industrial robot arms have been in use for industrial automation in automotive and other
manufacturing businesses for more than three or four decades. But various strands of
existing and newer research fields, such as AI and sensing, have been combined in more
recent years to produce autonomous “advanced” robots with more widespread potential use
across the economy and society.
1.1 What is a robot? Evolving definitions
In part driven by aforementioned Hollywood movies, most laypersons perceive “robots” to be
primarily, or exclusively, humanoid robots. However, humanoid robots are only a small
subset of the robotics industry.
Encyclopedia Britannica defines a robot as “any automatically operated machine that
replaces human effort.” According to the International Federation of Robotics (IFR), “[a]
robot is an actuated mechanism programmable in two or more axes with a degree of
autonomy, moving within its environment, to perform intended tasks”.2 In turn, the majority
of practitioners and scholars consider a robot to be any “machine capable of sensing its
environment and reacting to that environment based on an independent decision-making
The term autonomy is often used to underline the difference between robots and other
machines; a robot has the ability to interpret its environment and adjust its actions to achieve
a goal. In terms of technological trajectory, robots are evolving from programmed
automation, over semi-autonomous to more autonomous complex systems. Fully-
autonomous systems are able to operate and make “decisions” to complete tasks without
human interaction (see Box 1 for a fuller definition).
Box 1 – Robotics evolution and definitions
A remote control device is a device that can be controlled from a remote location. Based on
the most common definition of a “robot”, remote controlled devices would not be considered
a “robot”. Nevertheless, the robotics industry has accepted certain purely remote controlled
devices as within the industry of robotics. For example, telepresence devices are often
referred to as robots, or even robotic-telepresence devices, despite the fact that some
telepresence devices are purely remote controlled. The same is true for certain toys, as well
as for certain educational devices. Examples of remote controlled devices frequently
considered to be within the bounds of the robotics industry include telepresence robots,
remotely controlled humanoid robots, robotic assisted surgical devices, exoskeletons, and
Unmanned Aerial Vehicles (UAV), also known as Unmanned Aerial Systems (UAS) or
Semi-autonomous devices still interact with and are controlled by human operators, but are
not purely remote controlled devices as semi-autonomous systems provide guidance to their
human operators to ease and/or assist in the device’s operation. Robots whose operation is
2 See IFR.
3 See Springer (2013), p. 1-5.
managed by a human operator is generally considered semi-autonomous. Examples
include semi-autonomous features frequently becoming more common in cars, as well as
certain industrial robots that require an operator to provide detailed commands.
Fully-autonomous devices are those devices that are able to operate, including the ability to
make “decisions” within the environment for which it was designed and in order to further the
task it was designed to accomplish, without human interaction. Fully-autonomous devices
are not typically designed to think creatively, but there is some admittedly blurred lines as
certain fully-autonomous devices are necessarily designed to operate in unpredictable
environments for which the fully-autonomous devices’ decisions are not predetermined.
AI is generally defined as its own area of computer science that is focused on computer-
based devices being capable of making intelligent human-like decisions. Although
admittedly a blurry line, one divide between fully-autonomous and AI is the difference
between a device making basic unsophisticated decisions (autonomy) and a device that
makes creative decisions. AI is considered by some practitioners to be within the robotics
industry, but many other practitioners consider artificial intelligence to be its own field of
technology; albeit with potentially profound implications on the robotics industry. The later
view is based on the understanding that artificial intelligence is grounded in computer-
science without necessarily any hardware application. Although artificial intelligence
incorporated into a movable hardware device is anticipated as a potential reality within the
robotics industry, artificial intelligence can be entirely separate from any hardware device.
1.2 History of Robotics
From industrial arms for automation
Robots, in their most basic form, are not new. The history of robotics started in ancient
Greek with automatons, essentially non-electronic moving machines which displayed moving
objects. The invention of simple automatons continually evolved henceforth, but robots in
their current form took off with the process of industrialization, to perform repetitive tasks.
In the more recent history of industrial robots, a few key inventions in two areas stand out as
having led to the first incarnation of robots for industrial automation.4 First, control systems
allowing humans or computers to control and steer robots from a distance, and second,
mechanical manipulation systems such as robotic arms or legs to move or grab objects.
As for mechanical manipulation systems, the first industrial robot was developed in 1937 in
the form of a small crane. The development of robotic legs and arms was furthered by W.G.
Walter, who built the first autonomous robot in the late 1940s.5 The breakthrough enabling
the development of the robotics industry, however, was when George Devol invented and
patented the first automatically operated programmable robotic arm in the mid-1950s.6 Devol
then partnered with Joseph Engelberger, considered by many scholars to be the “Father of
Robotics”,7 to create a company called Unimation, which produced a robot in 1956 based on
Devol’s patents. This started the commercialization of industrial robots.8
4 See IFR (2012).
5 US Patent 2,679,940. Willard L.V. Pollard and Harold A. Roselund, working for DeVilbiss Co., filed a patent for
the first programmable mechanized paint-sprayer in 1942.
6 US Patent 2,988,237. See also Nof (1999).
7 It should be noted, however, that many scholars and practitioners, especially those that consider remote
controlled devices to be part of the robotics field, also consider Nikola Tesla to be the “Father of Robotics”
Robotic arms have since been fine-tuned and improved. The first computer-controlled
revolute electric arm, for instance, was developed at the Case Institute of Technology, Case
Western Reserve University, US. In 1969, researchers at Stanford University invented the
so-called Programmable Universal Manipulation Arm, allowing for more sophisticated control
for assembly and automation.9 One of these researchers, Victor Scheinman, started Vicarm
Inc. to manufacture the arm, which proved fundamental to the development of the robotics
industry; he ultimately sold the company to Unimation in 1977.
Largely based on the work of the aforementioned inventors and firms, the first commercial
robots were deployed on General Motors’ assembly lines in the USA in 1961.10 The first
industrial robot in Europe, a Unimate, was installed in Sweden in 1967. In 1969, the
company Trallfa of Norway offered the first commercial painting robot. In 1973, ABB
Robotics and KUKA Robotics brought their first robots to market. Since then, the
functionality and control of robotic mechanical parts have been continually improved by the
Approximately a decade after Devol filed his patent, Japanese companies began to develop
and produce their own robots pursuant to a license agreement with Unimation. By 1970,
robotic manufacturing had proliferated throughout the automotive industry in the US and
Japan. By the late 1980s, Japan – led by the robotics divisions of Fanuc, Matsushita Electric
Industrial Company, Mitsubishi Group and Honda Motor Company – was the world leader in
the manufacture and use of industrial robots.
Parallel key inventions in the area of packaging robots – for instance, the Delta packaging
robot developed at the Federal Institute of Technology of Lausanne, yielding 28 patents –
modernized the packaging industry.
A full-scale humanoid robot developed at Waseda University in Japan laid the foundation for
follow-on innovation in the field, facilitating enhanced human–robot interaction relevant to
today’s consumer-oriented robot markets. While many historians have discussed evidence
of early pre-computer based use of “legs” for movement, the earliest breakthroughs
concerning machines that could walk via two or more legs occurred in the 1960s and 1970s.
Yet, such technology is not yet prevalent within commercialized products despite the
decades of research that have occurred since the early breakthroughs in this area.
…toward autonomous systems built on artificial intelligence and connectivity
In the journey toward more capable robots, researchers have since worked on increasing
autonomy and improving interaction between humans and robots. New materials and
innovations in various fields outside the robotics area such as artificial intelligence (AI),
mechatronics, navigation, sensing, object recognition and information processing are the
core technological developments furthering robotics today.11 The research has become
based, at least in part, on his 1898 invention and patent of a remote controlled boat (see United States Patent
8 See Rosheim (1994).
9 Scheinman (2015).
10 IFR (2012).
11 Kumaresan and Miyazaki (1999).
In particular, innovation in software and AI will be key technologies for next-generation
robots. This matters to help robots maneuver and circumvent obstacles. The seminal
breakthrough in developing algorithms instrumental for robotic path planning took place in
the mid-1980s and is credited to Randall Smith and Peter Cheeseman.12 The result of such
seminal research on the problem of Simultaneous Localization and Mapping (SLAM) led to
the development of SLAM algorithms, which many robotics companies still use as of the
date of this report, albeit with modifications tailored to the environment and purpose of their
specific robot. Algorithms are increasingly central to how robots make more complex
decisions, for instance, how home or service robots simulate emotions. Researchers are
currently working on software that will mimic the human brain, honing language and
Based on improved connectivity, sensors and processing power, robots are becoming
increasingly data-driven, and linked over more intelligent networks. As such, innovation is
increasingly about software and hardware integration and thus the delivery of so-called
integrated robotic and intelligent operational systems. On the application level, the
development of autonomous vehicles and drones is seen as an extension of robotics.
2. The Economic Contribution of Robotics
Robots already have a demonstrable and significant impact on how manufacturing takes
place. Since the start of industrial automation in the 1970s, the uptake of robots in
manufacturing has increased significantly. The industrial robot market was estimated to be
worth USD 29 billion in 2014, including the cost of software, peripherals and systems
engineering (see table 1).
Table 1: Different estimates of the robotics industry revenues
USD 29 billion (2014)
USD 33 billion (2017)
Global market for industrial robotics IFR (2014a)
EUR 50-62 billion (2020)
Global market for industrial robotics
USD 3.6 billion
Global market for service robots (of which
USD 1.7 billion for domestic use)
As illustrated by figure 1 (top), the number of robots sold is increasing, reaching about
230,000 units sold in 2014, up from about 70,000 in 1995, and projected to increase rapidly
in the next few years. Japan, US and Europe were the initial leaders in terms of market size.
Interestingly, the respective shares of various world regions in global robotics sales has
changed little, with Asia leading followed by Europe and North America, and rather small
volumes in South America and Africa. Yet within Asia, China has gone from no robots in
1995 to overtaking Japan to become the largest robot market. The Republic of Korea is now
the second biggest user of industrial robots in Asia.13
In terms of sectors, the automotive industry continues to be the main driver of automation,
followed by the electronics industries (see figure 1, bottom). Innovation will enable more
flexible and small-scale manufacturing.
12 Smith and Cheeseman (1986).
13 In terms of robotic density, as at 2014 the Republic of Korea had the highest robot density in the world, with
437 units per 10,000 persons employed in the manufacturing industry, followed by Japan (323) and Germany
(282). In comparison, China’s density was 30, Brazil’s 9 and India’s 2 (IFR, 2014a).
Figure 1: Worldwide shipments of industrial robots on the increase, led by Asia and
the automotive sector
Source: Authors based on IFR World Robotics Database, 2014.
Note: The regions as shown here follow the definition of the IFR.
A novel robotics field is the production and use of service robots in areas outside of
manufacturing. This category includes robots intended for “professional use” in agriculture,
mining, transport – including the large field of unmanned aerial and land vehicles, space and
sea exploration, unmanned surveillance, health, education and other fields.14
The total number of professional service robots reached USD 3.6 billion in 2014, projected to
lead the growth of upcoming robotic use.15 The largest markets are Japan, the Republic of
Korea, the US and Europe. The sectors leading their use are defense, logistics and health.
Surgical robot device markets, at USD 3.2 billion in 2014, are anticipated to reach USD 20
billion by 2021.16
In addition, robotics in personal and domestic applications, another novel robotics field, has
experienced strong global growth with relatively few mass-market products, for example
floor-cleaning robots, mowers, robots for education and assistive robots for the elderly.17
With small to non-existent sales volumes even in 2012 and 2013, the sale of these robot
types took off exponentially in 2014 and onwards.
A few consultancy reports have emphasized the wide range of savings generated through
advanced robotics in healthcare, manufacturing and services, producing high estimates of
the benefits to economic growth.18 But quantifying the productivity-enhancing contribution of
robots in definite terms is challenging.
Robots can increase labor productivity, reduce production cost and improve product quality.
In the service sector in particular, robots can also enable entirely new business models.
Service robots provide assistance to disabled people, mow lawns, but are also increasingly
deployed in service industries such as restaurants or hospitals.
In terms of welfare, robots help humans to avoid strenuous or dangerous work. They also
have the potential to contribute solutions to social challenges such as caring for the aging
population or achieving environmentally friendly transportation.
In part, the economic gains of robots are directly linked to substituting – and thus automating
– part of the currently employed workforce.19
On the one hand, more productive labor helps keep manufacturing firms competitive,
avoiding their relocation abroad and creating higher-wage jobs. The robotics industry has
been particularly focused over the past several years on alleviating fears that the increase in
robotics will lead to a decrease in available jobs. Indeed, the robotics industry has
conducted research and found evidence to support projections that increased employment
opportunities will follow from innovations and development in the robotics industry.20 Many
predict that the advancement and proliferation of robotics may lead to an increase in jobs
within certain high income nations as a result of manufacturing re-shoring (also called
14 See IFR (2014b).
15 IFR (2014b).
16 Wintergreen Research Inc. (2015).
17 IFR (2014b).
18 The McKinsey Global Institute estimates that the application of advanced robotics could generate a potential
economic boost of USD 1.7 trillion to USD 4.5 trillion a year by 2025, including more than up to USD 2.6 trillion in
value from healthcare uses (McKinsey Global Institute, 2013).
19 Metra Martech (2011), Miller and Atkinson (2013), Frey and Osborne (2013) and Brynjolfsson and McAfee
20 Metra Martech (2011).
manufacturing in-shoring) because manufacturing previously outsourced to nations with
cheaper labor will be performed in high income nations due to robotics reducing the cost of
manufacturing.21 For example, at a robotics-focused Xconomy conference in April 2013, a
well-known roboticist explained that, contrary to the past three decades, during which
companies moved their manufacturing to various low income nations based on their low
labor costs, such manufacturing will be re-shored to the United States and other high income
nations. Such predictions are premised on the decrease in manufacturing costs that will
result from industrial and other manufacturing related robotics.22 If such predictions are
accurate, it may be true that the proliferation of robotics will increase jobs and economic
growth in high income countries. However, at least in the manufacturing sector, the creation
of jobs in high income countries may be at the expense of jobs in middle and low income
countries if their primary attraction for their manufacturing industry is low labor costs.23
On the other hand, the use of robots is certain to eliminate both low-skilled but also some
types of higher-skilled jobs hitherto unaffected by automation. Yet, there are still multiple
sources that predict robotics, as a whole, will lead to a decrease in available jobs. Many of
these later reports focus only on the loss of current jobs24 and do not account for the
creation of new job types that may not exist today
On balance, the employment effect of robotics is currently uncertain.
In terms of overall economic benefits, another question is whether robotic innovation has
diffused to low- and middle-income countries already with meaningful impacts. The installed
base of robots outside a few high-income economies and a few exceptions such as China is
still limited, including in countries such as Brazil or India, but in particular also in less
developed economies. It is expected, though, that firms involved in manufacturing and
assembly activities for global or local supply chains will need to upgrade their use of robots,
including some in middle-income or even low-income economies that have so far competed
on cheap labor alone. Robots are also gaining ground in low-income countries to address
quality issues in local manufacturing.
Although the location of companies developing robots and robotics products appears to be
occurring in specific high income nations, the impact of certain robotics technologies may
have an impact on middle and low income countries akin to that of the internet. In the same
way that the internet has allowed certain jobs to be performed remotely, whether from a
location in the same region or a different continent, robotics technologies such as
telepresence robots or remotely controlled robots with arms will continue to increase the type
of jobs that can be performed from a remote location.25 As the internet and robotics
21 See Green, T., Robots, Re-Shoring and America’s Manufacturing Renaissance. Robotics Business Review,
June 10, 2012. See also Christensen et al. (2013), pp. 3 and 10: “The sale of robotics for manufacturing grew
44% during 2011, which is a clear indicator of the revitalization of the production system in the U.S. Robots have
been used as a facilitator to inshore manufacturing for companies such as Apple, Lenovo, Foxconn.”.
22 See Christensen et al (2013), p. 10 noting that the increased capabilities of manufacturing robotics and the
rising cost of labor in nations traditionally known for cheap manufacturing are both acting as incentives for
companies to inshoring manufacturing.
23 But see RBR Staff, China 2013: Factory Automation Driving Robot Growth, Robotics Business Review,
December 28, 2012 discussing Blue Paper by Morgan Stanley noting the low demand for robotics in China as of
2013 and predicting that China’s demand for robotics would increase due to China’s manufacturing industry’s
need to stay competitive.
24 See, as an example, Frey and Osborne (2013).
25 See Christiansen, et al. (2013), p. 66: “As the Internet continues to evolve, it will inspire a natural progression
from sensing at a distance to taking action at a distance. This extension of the Internet into the physical world will
serve to further blur the boundaries among community, communication, computing, and services and inspire new
dimensions in telecommuting and telepresence applications. Hybrid solutions are likely to emerge that enable
technology continue to evolve, making the location of an employee less important to
companies than the ability to perform certain tasks, any nation with sufficiently fast and
reliable internet service will be enabling its citizens to compete for jobs. In particular, middle
and low income countries may be able to compete for positions in higher income countries
for jobs requiring a higher level of intellect or creativity. This phenomenon already exists as
a result of the internet, with certain companies thriving on the offering of hiring individuals not
necessarily in a close geographic location.26 Such tasks are primarily restricted to purely
internet based deliverables, but with the advancement of certain robotics technologies, could
involve physical jobs as well. Although middle and low income nations may not benefit from
such advancements if they have slow, unreliable, or heavily restricted internet, certain
technologies are currently being developed to resolve such limitations.27
3. The Robotics Innovation System
As it evolves from the era of industrial automation to the use of advanced robotics across the
economy, the present-day robotics innovation system can be characterized by a few key
3.1 Concentration in key countries and narrow robotics clusters with strong
Robotics innovation mainly takes place within a few countries and clusters.28 These clusters
thrive on the interface between public and private research, with firms commercializing the
An analysis of robotics company databases shows that robotics clusters are mainly located
in the US, Europe – in particular Germany, France and to some extent the UK – and Japan,
but increasingly also in the Republic of Korea and China (see Figure 2).29 Applicants from
the same origins account for the vast majority of patent applications in the field of robotics
(see Figure 3). Relative to GDP or population size, Canada, Denmark, Finland, Italy, Israel,
the Netherlands, Norway, the Russian Federation, Spain, the UK, Sweden and Switzerland
stand out as economies with a big presence of innovative robotics firms.
distributed human cognition and enable the efficient use of human intelligence. Such solutions will combine the
robotics-enabled capability to remotely and autonomously perceive situations requiring intervention with the
Internet-enabled capability for human operators to take action from a distance on an as-needed only basis.”.
26 See, e.g., Sarah Halzack, Elance-oDesk flings open the doors to a massive digital workforce, The Washington
Post, June 13, 2014.
27 See Juliette Garside, Facebook Buys UK Maker of Solar-Powered Drones to Expand Internet, The Guardian,
March 28, 2014; See also Eliana Dockterman, Facebook Eyes Using Drones to Deliver Internet, Time Magazine,
March 27, 2014; Gregory S. McNeal, Google Wants Internet Broadcasting Drones, Plans To Run Tests In New
Mexico, Forbes, September 19, 2014.
28 Green, T., Rising Power and Influence of Robotics Clusters. Robotics Business Review, February 22, 2013.
29 Although there is no standard global database for all robotics companies, there have been some attempts to
compile such lists, see Tobe (2013).
Figure 2: Main geographic location of robotics companies, 2015
Source: Authors based on available information from robotics-focused associations and
groups, including The Robot Report’s Global Map as in Tobe (2015), as well as from the
publicly available listing of companies from the Robotics Industry Association (RIA).
Figure 3: Geographic distribution of robotics innovation, 1960-2011
Notes: Only first filings with at least one patent granted within the patent family
Source: WIPO based on the PATSTAT database (see technical annex).
This picture of inventive activity concentrated in a few nations, also now broadening to
include Asian innovative nations, is also mirrored by patent data. Patent data presented in
Section 4 attests the importance of US and European and later Japanese inventors at the
First filings by origin
1960 1970 1980 1990 2000 2010
JP CN KR US DE FR AU SE Other
outset, the emergence of the Republic of Korea in the early 2000s and more recently
Within these few countries, robotics clusters are concentrated around specific cities or
regions – and often around top universities in the field. For example, in the US, Boston,
Silicon Valley and Pittsburgh are generally regarded as the three main robotics clusters. In
Europe, the Île-de France region in France (particularly for civil drones), Munich in Germany,
Odense in Denmark, Zurich in Switzerland and Robotdalen in Sweden are prominent,
among others. In Asia, Bucheon in Korea, Osaka and Nagoya in Japan and Shanghai and
Liaoning Province in China are key robotics clusters.
Some companies that excel in robotics innovation are located outside these clusters. They
are usually established large companies in the automotive sector, or increasingly also
Internet companies, that are well-established in their own field. They have the financial
means and the skills to hire robotics experts and to use knowledge developed elsewhere,
also often by acquiring newer firms.
Indeed, in terms of robotics innovation and company startups, the majority of activity is in
high-income countries, except for China again. China has seen a strong surge of robotics
patents and hosts some of the fastest-growing robotics companies such as DJI (Drone
Company), and new industrial robot manufacturers such as Siasun and Estun which are
driving down the cost of robots.
3.2 Highly dynamic and research-intensive collaborative robotics innovation
The robotics innovation ecosystem comprises a tight and cooperative network of actors,
including individuals, research institutions and universities, and large and small technology-
intensive firms. Robotics brings together diverse science and technology breakthroughs to
create new applications; while long established, it continues to deliver new inventions as
new materials, motive power, control systems, sensing and cyber systems kick in.
As evidenced in section 1.2, individual entrepreneurs and their startups played a critical role
in kick-starting and further developing the robotics industry.
Select public research institutions are also crucial actors in the robotics innovation
ecosystem. Examples of leading universities include McGill in Canada, Carnegie Mellon in
the US, ETH in Switzerland, Imperial College in the UK, Sydney University in Australia,
Osaka University in Japan, and the Shanghai Jiao Tong University in China. PROs such as
the Korean Institute of Science and Technology, Fraunhofer in Germany, the Industrial
Technology Research Institute in Taiwan (Province of China) and the Russian Academy of
Sciences are notable too.
Traditionally, these science institutions play an important role in innovation generally by
conducting long-term research whose commercial applications will only be realized far in the
future.31 As depicted in figure 4, the role of academic and public institutions – as well as
those of individual entrepreneurs – varies through time and across countries.
In addition, however, in robotics specifically they had and continue to have a major role in
furthering development by creating spin-outs and spin-offs, by patenting (see section 4), and
30 See also UKIPO (2014).
31 See Nof (1999), p. 33 stating that the development of electric robots began in university labs.
through close collaboration with firms.32 Examples of spin-offs include Empire Robotics, a
spin-off of Cornell University, and Schaft Inc., a spin-off of the University of Tokyo.
Collaboration between firms and PROs is tight too, with, for instance, KUKA developing
lightweight robots with the German Institute of Robotics and Mechatronics. Furthermore,
their increased offering of formal robotics degrees has been critical in the development and
diffusion of skills, as corporations hire recent graduates.
Figure 4: First patent filings in the field of Robotics by origin and type of applicant,
Source: WIPO based on the PATSTAT database (see technical annex).
There are also examples of universities collaborating with the private sector beyond simply
accepting monetary support for robotics research, including, for example, universities
entering into joint development agreements with private companies for the purpose of
building robotic technology to solve problems relevant to the private company.33
Although universities may be more willing to focus on research and development that does
not necessarily have near term commercialization potential, it does not mean that
32 Nof (1999).
33 See RBR Staff, Mining Giant Anglo-American Inks Deal with Carnegie Mellon, Robotics Business Review,
January 9, 2013 discussing Carnegie Mellon University’s five-year joint development agreement with Anglo
American Plc, pursuant to which CMU’s Robotics Institute will design, build and deploy mining robots, robotics
tools and autonomous technologies in partnership with Anglo American’s internal Technology Development
Group. See also Deere & Company’s joint development with University of Illinois as evidenced in jointly assigned
United States Patent Nos. 7,587,081; 8,712,144; 8,737,720; and 8,855,405; and Deere & Company’s joint
development with Kansas State University as evidenced in the jointly assigned United States Patent Nos.
7,792,622; 7,684,916; 7,580,549; and 7,570,783; and MAKO Surgical Corp.’s joint development with the
University of Florida as evidenced in jointly assigned United States Patent Nos. D625,415 and D622,854.
1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010
1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010
CN JP KR
US DE FR
Academia and public sector Individuals Companies Not available
universities do not seek protection for their inventions, nor does it mean that such research
does not directly lead to new robotics products or robotics companies. Indeed, universities
conducting research in the field of robotics frequently patent inventions. Additionally, there
exist, as of the date of this report, numerous companies within the robotics industry that are
“spinoffs” or “spinouts” of research and development projects conducted at universities.34
When it comes to inventive robotics firms, three main types can be identified.
First, there are small company startups or specialized robotics firms which are often created
by individual inventors affiliated to academic robotics centers or robotics clusters, sometimes
with significant direct or indirect government support. An example is Universal Robots,
which emerged from a robotics cluster in Demark with links to the Danish Technological
Institute, receiving initial government and seed funding.
Although parts of the industry are more mature today, the potential for small robotics
startups is still large. In the early stage of radical innovation, small startups demonstrate
more agility and speed, and closer interaction with academia. Also, innovation ecosystems
are becoming more specialized, allowing for niche specialist companies. Third-party
external developers are increasingly part of the robotics innovation system, as robotics
platforms, often based on open-source software architectures, are the starting point for
further development. Also, a growing number of companies provide robotics-related
services – mobility or machine management systems. Moreover, the rise of new, more
consumer-oriented robotics firms and new funding mechanisms allow for small initial start-
ups. Play-i, now called Wonder Workshop, for instance, which focuses on creating
educational toy robots, recently raised money through crowd-funding platforms.
Second, large, established robotics companies, initially focused on industrial robot research
and production alone, such as ABB (Switzerland), Kawasaki Heavy Industries, Yaskawa and
Fanuc (Japan) and KUKA (Germany) are active in robotics R&D. Scale matters, as
innovating in the field of industrial robotics hardware is particularly capital-intensive; research
takes years to materialize. Large clients in the automotive sector, for instance, are only
willing to buy from large, trusted, established companies to avoid safety risks. In addition,
large robotics firms are emerging from the novel trend toward service and household robots.
iRobot (US) is one such example. Initially a spin-off from MIT, it is now a large company
producing robots for business, private households and for security purposes, although still
making most of its revenue from the development and sale of robots for military applications.
Third, large firms outside the robotics industry have also gained related competencies.
Firms such as BAE Systems (UK) in the area of defense, aerospace and security have
always and continue be important players for robotics innovation. In addition, firms in the
automotive sector continue to be significant, not least due to their own important use of
robots. A newer development is the increasing involvement of electronics and ICT firms
such as Samsung (Republic of Korea) and Dyson (UK). As robotics becomes more reliant
on connectivity and ICT networks, Internet or IT-related firms such as Amazon, Google and
Facebook but also the Indian ICT services firm Infosys, Alibaba of China and Foxconn of
Taiwan Province of China are joining the fray, often acquiring shares in or taking full
34 See Rory Cellan-Jones, Oxford’s robots and the funding of innovation, BBC.com, November 3, 2014. Startup
companies in the robotics field that are spinoffs/spinouts of research conducted in a university setting are
common in the United States, but also occur with some frequency in European and Asian nations. Examples
include Oxbotica (a spinout of Oxford University), Empire Robotics (a spinoff of Cornell University), Blue Belt
Technologies, Inc. (a spinoff of Carnegie Mellon University), Meka Robotics (a spinoff of Massachusetts Institute
of Technology and acquired in late 2013 by Google), Medrobotics Corp. (a spinoff of Carnegie Mellon University
and recently raised 20,000,000.00 USD in a Series F funding round), Schaft Inc. (a spinoff of University of Tokyo
and acquired in late 2013 by Google), RE2, Inc. (a spinoff of Carnegie Mellon University), and Autonomous
Solutions, Inc. (a spinoff of Utah State University).
ownership of established robotics firms. Moreover, firms in the health sector are also
increasingly prominent in robotics research. Market leaders in the area of surgical robots,
for instance, include Intuitive Surgical, Stryker and Hansen Medical.
As the advantages of robotics have become increasingly visible to non-robotics companies,
there has been an increased focus on traditional companies obtaining the advantages of
robotics capabilities applicable to their business. Such a desire to embrace robotics
technology has materialized into significant strategic business decisions, including,
acquisitions of robotics companies whose technology will likely directly benefit the acquiring
company’s business and/or risk replacing the acquiring company’s business,35 traditional
companies entering into joint development agreements with robotics companies for the
purpose of developing robotics solutions aimed at the traditional companies’ business,36
traditional companies creating their own internal robotics divisions through the hiring of
individuals with robotics experience,37 and the formation of strategic alliances for the
purpose of creating a new robotics ecosystem or “cluster”.
There are also recent examples of private companies attempting to tackle particularly difficult
problems in the robotics industry by using monetary incentives in a similar crowdsourcing
competition program akin to that of competitions organized by governments. For example,
in September 2014, Amazon.com Inc. announced the “Amazon Picking Challenge”, which
“challenges” company and university teams to solve the complicated problem of warehouse
picking.38 This type of “challenge” is similar to the open entry competitions known as the
United States’ DARPA Robotics Challenge or the prior DARPA Grand Challenge.
Both from publicly available information concerning formal partnerships and joint
development agreements, as well as from an analysis of patent filings covering robotics
technology,39 there is sufficient evidence to observe a significant degree of collaboration on
the development of robotics technology. There are several reasons why there is a greater
amount of collaboration in the robotics industry than in many other industries.
First, when it comes to government contracts, and in particular, defense contracts with the
United States Government, the contracted project is sometimes divided among more than
one robotics company. For example, the government will frequently award the design and
35 Notable examples of traditional companies acquiring robotics companies include (i) Amazon.com Inc.’s
acquisition of Kiva Systems, Inc., See John Letzing, Amazon Adds That Robotic Touch, The Wall Street Journal,
March 20, 2012 discussing Amazon.com Inc.’s acquisition of Kiva Systems, Inc.; (ii) Stryker Corp.’s acquisition of
MAKO Surgical Corp., See Joseph Walker, Tess Stynes, Stryker to Acquire Mako Surgical for About $1.65
Billion, The Wall Street Journal, September 25, 2013 discussing Striker’s acquisition of MAKO Surgical Corp.);
and (iii) Advantech Co.’s acquisition of LNC Technology, See Kevin Chen, Advantech buys majority stake in
LNC, Taipei Times, August 31, 2013 (discussing Advantech Co.’s acquisition of a majority of the shares of LNC
36 Notable examples of traditional companies entering into joint development agreements with robotics
companies for the purpose of developing robotics technology directly applicable to the traditional company’s
business include (i) Anglo American’s partnership with Autonomous Solutions, Inc.; (ii) Lowe’s Companies, Inc.’s
partnership with Fellow Robots; and (iii) John Deere’s joint development with iRobot Corporation, whose joint
inventions can be noted in jointly assigned patents (See United States Patent Nos. 8,874,300; 7,894,951;
8,020,657; and 8,473,140).
37 See George Anders, Amazon’s Drone Team Is Hiring: Look At These Nifty Job Ads, Forbes, May 19, 2014
discussing Amazon.com Inc.’s job postings seeking individuals with experience relevant to its then-recently
formed drone division.
38 Joe Romano, Amazon Picking Challenge, RoboHub, November 4, 2014; See also Merce Gamell,
Crowdsourcing for Designing the New Amazon Robots, ARISPlex, November 20, 2014.
39 See UKIPO (2014), p. 13-16 observing that the data obtained from patent filings corroborates the conclusion
that there is a substantial amount of collaboration within the robotics industry.
fabrication of the mechanical and electrical aspects to one company, but require the
software to be designed and built by a different robotics company.40
Second, the problems that robotics companies tackle are often extremely complex in
multiple disciplines. As a practical matter, most small and midsize robotics companies do
not have experts in all of the engineering disciplines necessary to build all aspects of a
sophisticated robot. The complexity of the technological challenges applicable to many of
the robotics products being developed is one reason that even well-capitalized robotics
companies enter into joint development agreements with other robotics companies.41
Third, customized autonomous systems are now a common site in medical device
companies, pharmaceutical companies, and laboratories. Some companies and labs have
an internal robotics and automation group that will work on certain projects independently,
but who will also collaborate with specialized robotics companies when presented with a
particularly time consuming or otherwise challenging assignment.42
The high degree of collaboration surrounding the development of robotics technology
suggests that joint development projects, as well as personal contacts, are both meaningful
mechanisms and avenues through which knowledge and skills relevant to robotics
technology is diffused. Although the degree to which such knowledge and skills are also
diffused through scientific publications and the publications of patent applications is
unknown, this report’s author is confident that both types of publications are a mechanism
used within the robotics industry for acquiring knowledge of technological developments
within the industry.43
40 See, as an example, Robotics Trends Staff, Autonomous Solutions Awarded Contract to Develop an
Immersive UI, Robotic Trends, July 5, 2007.
41 Notable examples of joint development agreement between robotics companies includes iRobot Corporation’s
joint development and licensing agreement with InTouch Technologies Inc. d/b/a InTouch Health, See iRobot
Press Release, iRobot and InTouch Health Announce Agreement, July 20, 2011; See also InTouch Health Press
Release, iRobot and InTouch Health Announce Agreement To Explore Potential Opportunities in Healthcare
Market, July 20, 2011; See also, as evidence of the joint development between iRobot and InTouch, jointly
assigned United States Patent No. 8,718,837.
42 Although internal robotics and automation divisions within traditional companies are not necessarily publicly
promoted, a search of robotics-focused patents assigned to traditional companies outside the technology industry
demonstrate that robotics related innovation is at least occurring within traditional companies in areas relevant to
those companies’ business. See, as examples, robotics and automation related patents assigned to Pfizer Inc;
see, e.g., United States Patent Nos. 5,370,754; 6,489,094); Abbott Laboratories (see, e.g., United States Patent
Nos. 6,588,625; 8,318,499; and Deere & Company. see, e.g., United States Patent Nos. 7,861,794; 8,195,342;
8,295,979; and 8,874,261.
43 As part of the qualitative research conducted for this report, C. A. Keisner conducted informal interviews with
directors of IP for numerous prominent robotics companies. Among the information collected from such informal
interviews, was that several robotics companies regularly monitor the patent applications published by their
competitors. Regularly monitoring the patent applications of competitors at the time such applications are
published is perceived by those doing such monitoring to serve three primary purposes, including, (i) learn of new
developments in technology relevant to their business, (ii) obtain insight into a competitor’s plans to either
improve an existing product or create a new product, and (iii) learn if a competitor is attempting to obtain patent
protection concerning something that should be challenged as either non-novel or obvious. However, it is not
merely competing robotics companies who monitor the publication of patent applications in order to be aware of
technological developments in the robotics industry. Indeed, when a patent application is published concerning
an invention of particular interest, especially when it relates to a potentially transformative technology and/or
indicates a potentially shift in direction for a company well-known to the public, it is not uncommon for the patent
application to become the subject of an article. See, e.g., Jason Falconer, Patent Suggests Sony Still Sees
Future for Household Robots, IEEE Spectrum, April 14, 2014 discussing Sony Corporation’s potential strategy to
develop personal robots based on the publication of United States Patent Application No. 2014/0074292, filed on
April 16, 2012, entitled “Robot device, method of controlling robotic device, computer program, and program
Generally speaking, the exchange of knowledge within the robotics ecosystem currently
seems extensive and fluid. This is benefited by the science-intensive nature of robotics
innovation and the strong role of science and research institutions, but also the admittedly
nascent phase of many advanced robotics strands. Scientific papers and conferences –
such as the International Symposium on Industrial Robots – play a key role in the transfer of
knowledge. Moreover, robotics contests and prizes rewarding solutions to specific
challenges enable researchers to learn and benchmark their progress, and to close the gap
between robotics supply and demand. Collaboration among the three types of firms
mentioned above is extensive.
Finally, decentralized, software-enabled innovation is likely to increase in the future as
robots become more widespread, and robot platforms and systems more standardized. In
practice, a wider set of external firms and partners will be able to deliver customized
solutions to existing proprietary robotic software platforms. This will enable greater
modularity in innovation.
3.3 The extensive role of government in orchestrating and funding innovation
Governments and their institutions have played a large role in supporting robotics innovation.
The standard set of technology-neutral government innovation policies has strongly
supported robotics innovation, in particular through supply-side policies taking the form of
research funding or support for business R&D.
In particular government-funded military technology advancements that, while kept
confidential for years, are ultimately disclosed and contribute to the progress of the robotics
industry.44 Advanced technologies, although originally developed for military applications,
are often utilized by the private sector for non-military commercial purposes.
Beyond important research funding and standard innovation support measures, a few
specific support measures deserve mention:
Creation of special research institutions or research networks: Examples include the
Swiss National Centre of Competence in Research Robotics, which federated research labs,
and the Korea Robot Industry Promotion Institute, set up to promote technology transfer.
R&D funding, grants and public procurement: Governments, and often the military, fund
robotics innovation and create demand by the means of grants or – often pre-commercial –
procurement. In the US, R&D contracts, including from the National Institutes of Health or
DARPA, are the foremost catalysts.45 Pre-commercial procurement of robotics solutions for
the healthcare sector, for instance, is part of EU Horizon 2020 grants. Governments have
also incentivized innovation and advancement within the robotics industry through various
types of incentive programs. In the United States, for example, the government incentivized
private companies and universities to create autonomous vehicles by offering a substantial
monetary sum to whomever accomplished a set task.46 Other governments have provided
tax breaks for robotics companies, although some may argue that such incentives primarily
incentivize the relocation of robotics companies rather than fostering innovation. Yet, still
other countries, whether focused specifically on robotics or not, provide grants allowing
prototype-stage products to be used within an industry with potential customers, thereby
44 See Springer (2013), p. 15-16 discussing the development and use of drones for military purposes in the
45 Mireles (2006), Springer (2013) and Siegwart (2015).
46 See Mireles (2006) discussing the incentive scheme created by the United States’ Department of Defense’s
research arm, the Defense Advanced Research Projects Agency (DARPA), with the goal of furthering the
advancement of driverless vehicles.
alleviating the particularly lengthy time gap for robotics companies between the creation of a
functional prototype and a commercialized product that has been more rigorously tested and
satisfied arduous regulatory requirements.47
Organizer of contests and challenges and prizes: Governments have played a role as
organizer of robotics contests. Japan has announced a Robot Olympics, the UK recently
held a competition for driverless vehicles and the DARPA Robotics Challenge is a landmark.
Incentives for collaboration, technology transfer, finance and incubation: Through
grants or contracts, governments will frequently require collaboration and technology
transfer. The EU Horizon 2020 Robotics project, for instance, stimulates public-private
collaborative projects of a multi-disciplinary nature. In addition, government activities aim to
facilitate cluster development, entrepreneurship and industry networking. Governments also
ease the financing of robotics innovation, for example, the French government’s seed fund
Regulations and standards: Finally, regulations created by governments, in the form of
standards, testing and security regulations, impact the diffusion of robotics technology.
Legal scholars disagree about the extent to which regulations actual spur or inhibit the
growth of technological advancement in the robotics industry.48 There does seem to be a
general consensus that regulations have the potential to restrict such advancements.49 One
of the primary areas in need of attention is the reform of current safety standards applicable
to robotics, particularly those requiring the clear separation of workspace between humans
and robots.50 The absence of reform to such safety standards, as may apply to other
regulation, may hinder innovation and adoption of robotics technology.51 Furthermore,
governments can also hinder innovation in the private sector via burdensome regulations.52
Yet, aside from the restricting regulation applicable to UAVs in several nations,53 there is no
regulation specific to most other robotics related technologies.54
In addition to the above, many high-income countries and China have announced special
robotics action plans in recent years (see table 2). Mostly, these plans announce specific
47 See Technopolis and University of Manchester (2011).
48 See RoboLaw, Guidelines Regulating Robotics, p. 10 (2014) discussing the contradicting perceptions that
premature and obtrusive legislation might hamper scientific advancement, with the perception that a lack of a
“reliable and secure legal environment” may equally hinder technological innovation.
49 See RoboLaw, Guidelines Regulating Robotics (2014); See also, e.g., Ed Pilkington, What’s keeping America’s
private drone industry grounded?, The Guardian, September 30, 2014 discussing the impact of regulation on the
United States’ drone industry.
50 See Christensen, et al. (2013), p. 84 noting that “safety is a multidimensional issue extending beyond
technology. It includes a number of governmental and industry standards as well as independent certification
and liability exposure. These non-technical elements need to progress such that that clear standards exist for
both professional and personal robotics providing all stakeholders with visibility needed for rapid innovation and
51 See Christensen, et al. (2013), p. 84 noting that the “current, limited set of standards for safety certification of
both professional and personal robots, constrains innovation, reduces the pace of adoption, and adds costs.”.
52 See Infra, Section II.B; See also Free the Drones, The Economist, December 6, 2014; RoboLaw, Guidelines
Regulating Robotics, p. 10 (2014).
53 Regulations concerning UAVs exists in at least Canada, Australia, the United States, and the European
nations whose airspace is regulated by the European Aviation Safety Agency.
54 See Report of the Sixty-Eighth Session of the Working Party on Road Traffic Safety, Economic and Social
Council, United Nations, ECE/Trans/WP.1/145, April 17, 2014 (proposing amendments to the 1968 Vienna
Convention on Road Traffic in consideration of driverless/autonomous vehicle technology); See also Legislation
in the United States concerning driverless cars, including enacted state legislation in California (SB 1298), District
of Columbia (B19-0931), Florida (CS/HB1207), Michigan (SB 0169, 0663), and Nevada (AB 511, SB 140).
monetary investments in support of robotics research and innovation, including improving
robotics education and technology transfer.
Table 2: National robotics initiatives
Name of initiative
Country (Year of initiative)
National Robotics Initiative Advanced Manufacturing
France Robots Initiatives/
Feuille de Route du Plan
Robotics project Horizon 2020
New Industrial Revolution Driven by Robots (“Robot
Next-Gen Industrial Robotization
Republic of Korea (2015)
Robotics technology roadmap in 13th Five-Year Plan
Source: Authors based on national sources.
4. Robotics Innovation and Intellectual Property
The focus of robotics innovation is shifting from industrial automation to more advanced
robotics involving various technological fields, actors and economic sectors. As a result,
related IP and other strategies to appropriate returns on innovation investment are
embryonic; our understanding of them is incomplete. Also, recognizing the broad scope of
the robotics industry is important for this study because the large variety of robotics products
and their applications means that there is no “one-size-fits-all” IP strategy for robotics
companies, nor are observations and trends related to one segment of the robotics industry
necessarily relevant to other segments of the robotics industry.
Some tentative findings on appropriation strategies do, however, emerge on the basis of the
existing literature, data and insights from industry practitioners and robotics researchers.
4.1 The increasing role of patents, their function and potential challenges
Patents are a particularly important IP right for robotics companies because of the significant
amount of capital frequently required for research and development prior to the
manufacturing of a commercially ready product.
Indeed, the large pre-market research and development costs coupled with slow regulatory
clearance can create a context in which trailblazing robotics companies feel required to turn
to patent protection in order to recoup their investment. Absent this protection, newcomers
would be able to enter the market, after the “trail has already been blazed”, at a lower cost
for research and development and have to overcome fewer regulatory hurdles.55
For inventions discoverable through reverse engineering or other legal means, patent
protection is typically favored over trade secrets. It is understood that many robotics
companies whose competitive advantage is perceived to be sophisticated software designed
to enable robotic hardware devices may use software that is so complicated it cannot be
easily reverse-engineered, which is something traditionally believed to be possible with
software-based electro-mechanical devices.
55 See Cooper (2013), citing Casey Nobile & C. Andrew Keisner, The IP Battle Continues for Robotics
Companies: A Patent Attorney’s Reprise of the VGo/InTouch Health Verdict and its Implications, Robotics
Business Review, January 7, 2013.
Although deterring and excluding competitors is frequently a primary consideration of
robotics startups, another common incentive for seeking patent protection concerns the
perceived advantages to startups when seeking investments.56
As a result, key robotics inventions were frequently patented by their original – often
academic – inventor, who often also started a corresponding company or actively transferred
the IP to existing manufacturing firms.
In sum, robotics patents increased strongly during the 1980s, as broad-based automation of
factories flourished and robotics research was ramped up. Robotics-related first filings
roughly quadrupled during this decade (see figure 3). More importantly, these filings
outpaced patent filings from other technological fields. Robotics share of total patents
increased from .13% in 1980 to 0.53% in 1993 (see figure 5). Then, after a relatively flat
patenting activity during the 1990s and first half of 2000s, the shift to more advanced
robotics has given another boost to robotics patenting which continues to this day. In a
period of increasing overall patenting activity, robotics absolute patent filings roughly
doubled and the share increased from .4% in 2004 to .6% in 2011 (see figure 5).
Figure 5: Fast growth in robotics patenting, especially in the late 1980s and as of
First patent filings by origin, 1960-2011, in percent of all filings
Source: WIPO based on the PATSTAT database (see technical annex).
This picture of inventive activity concentrated in a few nations, also now broadening to
include Asian innovative nations, is also mirrored by patent data. Earlier, Figure 3 depicted
the number of first patent filings worldwide in the robotics space between 1960 and 2012. It
shows the importance of US and European and later Japanese inventors at the outset, the
emergence of the Republic of Korea in the early 2000s and more recently China.57 While the
56 See Keisner (2012); See also Eilene Zimmerman, Why More Start-Ups Are Sharing Ideas Without Legal
Protection, The New York Times, July 2, 2014 mentioning that early-stage investors are generally reluctant to
sign non-disclosure agreements and discussing the benefits of filing provisional patent applications in that
context. This motivation for startups to file patent applications was corroborated by several directors of IP for
robotics startups during informal interviews conducted by this report’s author in connection with this report.
57 See also UKIPO (2014).
as percent of all filings
1960 1970 1980 1990 2000 2010
share of Chinese patents in total robotics patents in 2000 was only two percent, that figure
had risen to 37 percent by 2011. The Republic of Korea’s share stood at 17 percent in 2011.
Japan’s share fell from 45 percent in 2000 to 10 percent in 2011.58
Figure 6 below indicates the origin of first patent filers in two periods, 1980-1990 and 2000-
2012. In the more recent period, the countries with the highest number of filings are Japan,
China, Republic of Korea and the US, which each filed more than 10,000 patents and
together account for about 75 percent of robotics patents, followed by Germany with roughly
9,000 patents and France with over 1,500. Other countries such as Australia, Brazil, a
number of Eastern European countries, the Russian Federation and South Africa also show
newer robotics patenting activity, although on a low level.
Indeed, in terms of robotics innovation and company startups, the majority of activity is in
high-income countries, except for China again.
Figure 6: Increasing but limited geographical diversity in robotics innovation
Number of first patent filings worldwide, 1980-1990
Number of first patent filings worldwide, 2000-2012
Source: WIPO based on the PATSTAT database (see technical annex).
58 Note that proportions are calculated considering only first filings with at least one patent granted within the
R b ti b fi t li t i i i d (1980 1990)
R b ti b fi t li t i i i d (2000 2012)
Actual robotics patent exclusivity is geographically highly concentrated, namely in Japan as
the leading destination with close to 39 percent of all robotics patents, the US and China with
close to 37 percent, Germany with 29 percent, followed by other major European countries
and the Republic of Korea. In turn, only 1.4 percent of robotics patents are filed on average
in other low- and middle-income countries.
Figure 7: Robotics patenting focused on a few selected destinations only
Share of robotics-related patent families worldwide for which applicants have sought
protection in a given country
Source: WIPO based on the PATSTAT database (see technical annex).
Automotive and electronics companies are still the largest filers of patents relating to robotics
(see table 3), but new actors are emerging from different countries and sectors such as
medical technologies. These firms’ robotics patent portfolios are growing in size, as firms
grow them organically or purchase companies with a stock of granted patents.
Table 3: Top 10 robotics patent filers, 1995-onwards
Company name Country
Number of first patent
Republic of Korea
Source: WIPO based on the PATSTAT database (see technical annex).
The large and growing stock of patents owned by universities and PROs is noteworthy too.
Table 4 lists the most important patent holders, now largely dominated by Chinese
universities. While industry experts note a strong move towards “open source” in the young
generation of roboticists at universities, the IP portfolios of universities are also growing
strongly, possibly facilitating the commercialization of new technologies as described in
earlier sections, but possibly also creating new challenges for universities and PROs in
managing and utilizing these sizeable portfolios.
Patent families (%)
60% or more
less than 1%
Table 4: Top 10 robotics patent holders among universities and PROs, 1995-onwards
Top 10 patenting worldwide
Shanghai Jiao Tong University
Chinese Academy of Sciences
Korea Institute of Science and Technology (KIST)
Rep. of Korea
Electronics and Telecommunications Research Institute
(ETRI) 289 Rep. of Korea
Tsinghua University 258 China
Harbin Engineering University
National Aerospace Laboratory
Harbin Institute of Technology
KAIST 188 Rep. of Korea
Top 10 patenting worldwide (excluding China)
Korea Institute of Science and Technology (KIST) 290 Rep. of Korea
Electronics and Telecommunications Research Institute
(ETRI) 289 Rep. of Korea
National Aerospace Laboratory (JAXA) 220 Japan
KAIST 188 Rep. of Korea
Deutsche Zentrum für Luft- und Raumfahrt 141 Germany
Fraunhofer-Gesellschaft zur Förderung der angewandten
Forschung 91 Germany
University of Korea 85 Rep. of Korea
Hanyang University 84 Rep. of Korea
Seoul National University 77 Rep. of Korea
National Institute of Advanced Industrial Science and
Technology (AIST) 69 Japan
Note: Depending on the legislation and policy in place, academic inventors may file under
their own name or the spin-off company name in certain countries (see WIPO, 2011). They
are not captured here.
Source: WIPO based on the PATSTAT database (see technical annex).
It is challenging to understand the various factors leading firms in the field of robotics to file
for patents, given the current evidence base. No large-scale survey of robotics firms or other
solid quantitative work exists that would shed light on this question. Providing a definitive
answer on the impacts of robotics patents on follow-on innovation via disclosure, licensing
and IP-based collaboration is also difficult.
However, a number of findings emerge from the views of industry experts, including both
lawyers and roboticists.
As in other high-tech sectors, and in anticipation of significant commercial gains from the
robotics industry, robotics firms seek to use patents to exclude third parties, to secure their
freedom to operate, to license and cross-license technologies and, to a lesser extent, to
avoid litigation. For small and specialized robotics firms in particular, patents are a tool to
seek investment or a means of protecting their IP assets defensively against other, often
In terms of the impacts of the patent system on innovation, at present the innovation system
appears relatively fertile. Collaboration – including university–industry interaction – is strong,
and there is extensive cross-fertilization of research. Patents seemingly help support the
specialization of firms, which is important for the evolution of the robotics innovation system.
It is also hard to argue that patent protection is preventing market entry or restricting robotics
innovation more generally by limiting access to technology. The available evidence shows
little or no litigation occurring in the field of robotics. Indeed, most of the disputes over
robotics IP in the past 10 years have involved just one company, iRobot.
The importance of particular patents for robotics innovation is hard to verify too. Currently,
no patents have been flagged as standard-essential; no known patent pools exist in the area
of robotics. And there are few formal and disclosed collaborations or exchanges in which IP
is central. Only one major licensing deal in the recent history of robotics has received much
attention.59 That said, company acquisitions involving the transfer of IP are growing
As regarding disclosure, firms use patents to learn of new technology developments, to gain
insight into competitors’ plans to improve or create products, but also to learn if a competitor
is attempting to obtain patent protection that should be challenged. Forward patent citations
within and outside robotics are often used as a sign that incremental innovation taking place;
earlier inventions are built upon. Often, however, and in particular in the US patent system,
they are a mere legal obligation, making impact assessment more difficult. As a result, the
overall value of patent disclosure in the area of robotics remains largely unassessed.
Many of the above questions will only be resolved over time. Arguably, IP is not yet fully
used in advanced robotics and so its potential impact remains to be realized. Compared
with the standard industrial robot innovation of the past, today’s robotic innovation system
involves more actors, various technology fields and significantly more patent filings. One
can start to see the more intensive offensive and defensive IP strategies that are present in
other high-technology fields.
A vital question is whether the increased stakes and commercial opportunity across various
sectors will tilt the balance toward costly litigation, as in other high-tech and complex
technologies. For the moment the number of IP disputes involving robotics companies is too
small to extract any meaningful insight about the effectiveness of the IP system. One
noticeable trend is also that the majority of IP disputes over the past ten years involved a
single well-known United States based robotics company, including a 2005 lawsuit captioned
iRobot Corp. v. Koolatron & Urus Indus.¸60 a 2007 lawsuit captioned iRobot Corp. v. Robotic
59 The most prominent agreement in recent history was the July 2011 joint development and cross-licensing deal
between iRobot Corp and InTouch Technologies.
60 In 2005, iRobot sued a Canadian company, Koolatron & Urus Indus., over its vacuum robot. iRobot asserted
several different types of IP claims including patent infringement, copyright infringment, and infringement of the
Roomba’s trade dress. iRobots’ Roomba, at the time of this lawsuit, was a round floor vacuum in grey and white,
whereas Koolatron’s vacuum robot was also round, but with yellow and black color. See iRobot Corporation v.
Urus Industrial Corporation, et al., 1:05-cv-10914 (D. Mass. 2005). The patents that iRobot alleged to be infringed
by Koolatron & Urus Indus. were United States Patent Nos. 6,594,844 and 6,883,201. This case quickly settled,
but the parties entered into a consent judgment with three main points. First, it said iRobot’s patents are valid
and enforceable. This consent judgment may not have an impact on other vacuum robot companies, but it
probably prevents Koolatron from ever arguing otherwise. Second, Koolatron agreed not to sell its vacuum robot
in the United States. And Finally, Koolatron agreed to not advertise in any way that dilutes iRobot’s reputation
FX,61 a 2013 lawsuit in Germany captioned iRobot v. Elektrogeräte Solac Vertrieb GmbH,
Electrodomésticos Solac S.A, Celaya, Emparanza y Galdos Internacional S.A, and Pardus
GmbH,62 and another 2013 lawsuit in Germany captioned iRobot v. Shenshen Silver Star.63
There was also only one recent IP dispute in which the lawsuit went to a final judgment and
appeal,64 which makes it difficult, if not impossible, to assess whether the judicial system
adequately resolves IP disputes involving robotics related technologies.
There have also been cases – though not many to date – in which non-practicing entities
have targeted robotics companies with a lawsuit.65 In particular, press reports mention the
or products. See Order Entering Consent Judgment, dated August 10, 2005, filed as Dkt. No. 2 in iRobot
Corporation v. Urus Indus. Corp., et al., 1:05-cv-10914 (D. Mass. 2005).
61 In 2007, a former iRobot employee started a new company called Robotic FX, which created a robot for
military mine detection that was similar to the robot he developed while he was an employee at iRobot. After
Robotic FX won a lucrative government contract by slightly underbidding iRobot, iRobot filed two lawsuits against
Robotic FX and its founder for trade secret misappropriation and patent infringement. See iRobot Corporation v.
Jameel Ahed and Robotic FX, Inc., 1:07-cv-11611 (D. Mass. 2007) (lawsuit filed in the United States District
Court for the District of Massachusetts by iRobot against Robotic FX and its founder for trade secret
misappropriation); iRobot Corporation v. Robotic FX, Inc., 2:07-cv-01511 (N.D. Ala. 2007) (lawsuit filed in the
United States District Court for the Northern District of Alabama by iRobot against Robotic FX for infringement of
United States Patent Nos. 6,263,989 and 6,431,296). The patents asserted by iRobot covered tread wheels in
the front that swiveled, which was used on both robots. This case was an extremely public and high profile case,
but it ultimately settled with some harsh consequences for Robotic FX and its founder, which included liquidating
Robotic FX and its founder agreeing not to work in the industry for five years. See Order Entering Stipulated
Consent Judgment and Proposed Permanent Injunction, dated December 21, 2007, filed as Dkt No. 88 in iRobot
Corporation v Jameel Ahed and Robotic FX, Inc., 1:07-cv-11611 (D. Mass. 2007). Additionally, the lucrative
contract that Robotic FX seemingly won by underbidding iRobot, did in the end, go to iRobot.
62 In July of 2013, iRobot filed a patent infringement lawsuit in the District Court of Düsseldorf against
Elektrogeräte Solac Vertrieb GmbH and three other companies, alleging that their vacuum robot, the Solac
Ecogenic AA3400, infringed upon five iRobot owned European patents. See Hiawatha Bray, Patent lawsuit
targets four iRobot rivals, The Boston Globe, July 2, 2013. The patents that iRobot asserted in this lawsuit were
EP 1 331 537 B1, EP 2 251 757 B1, EP 1 969 438 B1, EP 1 395 888 B1 and EP 1 776 623 B1. Id.
63 In September 2013, iRobot filed a patent infringement lawsuit in the District Court of Düsseldorf against
Chinese-based companies Shenzhen Silver Star Intelligent Technology Co. Ltd. and Shenzhen Silver Star
Intelligent Electronic Ltd. iRobot alleged that the Shenzhen Silver Star companies’ vacuum robot infringed on
four of iRobot’s European patents. Since iRobot was able to obtain a preliminary injunction against the
Shenzhen Silver Star companies based on its allegations of patent infringement, German [customs] officials
seized the robotic vacuums that the Shenzhen Silver Star companies were presenting at the 2013 IFA consumer
electronics trade show in Berlin, Germany. See RBR Staff (2013); See also Patricia Resende, Robot parts
seized at trade show after iRobot takes legal action against Chinese firm, BizJournals.com, September 6, 2013
(article including picture of customs officials seizing robotic floor vacuums from IFA consumer electronics trade
show); Mike Davin, Robots taken from trade show after iRobot obtains injunction, TheBusinessOfRobotics.com,
September 6, 2013.
64 InTouch Health and VGo Communications both independently developed telepresence robots and did not
appear to clash until VGo started to use its telepresence robot in hospitals. InTouch’s telepresence robot is
designed for hospitals and has specific features for doctors that VGo’s telepresence robot does not. But VGo’s
telepresence robot is a fraction of the price. Regardless of whether these robots are actually in competition for
the same potential customers, InTouch sued VGo on [date] for patent infringement and on [date] the jury came
back with a full victory for VGo, finding the patents were both not infringed and invalid. See Nobile (2013). The
Court of Appeal for the Federal Circuit affirmed the jury’s decision on non-infringement, but reversed on the issue
of the patent’s validity. See InTouch Technologies, Inc. d/b/a InTouch Health v. VGo Communications, Inc., 751
F.3d 1327 (Fed. Cir. 2014). It is important to note the significance of this decision in light of the growing number
of telepresence robots that are already on the market or soon will be. See Nobile (2013) discussing ramifications
of the jury verdict in the InTouch Technologies, Inc. v. VGo Communications lawsuit as it pertains to other
telepresence robotics companies. For such companies, decisions like this are important to their future both in
terms of future markets and their ability to obtain funding.
65 See e.g., Roy-G-Biv. Corp. v. Fanuc Ltd., 07-cv-00418 (E.D. Tx. 2007) (asserting infringement of United States
Patent Nos. 5,691,897; 6,513,058; 6,516,236; and 6,941,543); see also Roy-G-Biv Corp. v. ABB, Ltd., et al, 11-
cv-00622 (E.D. Tx. 2011) (asserting infringement of United States Patent Nos. 6,513,058 and 6,516,236); Roy-G-
Biv Corp. v. Honeywell Int’l, Inc., et al., 11-cv-00623 (E.D. Tx. 2011) (asserting infringement of United States
Patent Nos. 6,513,058; 6,516,236; and 8,027,349); Roy-G-Biv Corp. v. Siemens Corp., 11-cv-00624 (E.D. Tx.
possibility of negatively perceived patent troll activity in the field of surgical robots and
medical robotics more broadly.
Two elements could increase the likelihood of disputes. First, experts consulted in the
course of research for this report have raised concerns that overly broad claims are being
made in the case of robotics patents, especially with respect to older patents. While patent
infringement disputes between robotics companies appear to be resolved effectively by
current judicial systems,66 certain patent infringement disputes have led some professionals
within the robotics industry to question the breadth of patent claims contained in older
Second, in certain countries the patentability and novelty of computer-related inventions
generally are a matter of debate. This is particularly true in the US, where the recent
Supreme Court decision in Alice Corp. v. CLS Bank seems to have reinforced a restrictive
approach on the patent eligibility of software.68 Given the large and growing software-
related component of robotics innovation, concerns about software patentability may pose a
challenge in relation to current and future robotics-related patents.
4.2 Robotics platforms and the coexistence of IP and open source
As described in section 3 robotics platforms used in universities and businesses are
increasingly central to robotics innovation. Increasingly, too, they are open platforms, often
based on open-source software such as the Robot Operation System (ROS). These open-
source robotics platforms invite third parties to use and/or improve existing content without
the formal negotiation or registration of IP rights. Instead, software or designs are distributed
under Creative Commons or GNU General Public License, a free software license. This
allows for rapid prototyping and flexible experimentation.
The idea is simple. Actors distinguish between two levels of innovation. On the one hand,
there is the collaborative development of robotics software, platforms and innovation. Such
innovation may be substantial, but it is essentially precompetitive because the fields of use
are relatively basic and do not serve to differentiate products. Actors therefore apply
cooperative open-source approaches to obtain common robotics platforms, as this allows
them to share the substantial up-front investment, avoid duplication of effort and perfect
2011) (asserting infringement of United States Patent Nos. 6,513,058 and 6,516,236). For the purpose of the
discussion in this report, patent infringement lawsuits against robotics companies that concern technology not
specific to robotics are not considered. See, e.g., Hawk Technology Sys., LLC v. Fanuc Robotics Corp., 13-cv-
15234 (E.D. Mich. 2013) (asserting infringement of United States Patent Nos. RE43,462 and 5,265,410 covering
video recording and storage technology); Landmark Technology, LLC v. iRobot Corporation, 13-cv-00411 (E.D.
Tx. 2013) (asserting infringement of United States Patent Nos. 5,576,951 and 7,010,508 concerning internet
based e-commerce technology).
66 It should be noted that, for the purpose of this report, effective resolution of an IP dispute does not take into
consideration the resources expended by companies in asserting or defending against a claim, which varies
greatly amongst countries based on differences in judicial systems, including, for example, the large difference in
resources expended in a lawsuit based on discovery rules and whether there exist fee-shifting rules.
67 See, supra, discussion of InTouch Technologies v. VGo Communications; See also Frank Tobe, The patent
grip loosens, Everything-Robotics, December 6, 2012 discussing aggressive patent strategies by InTouch and
Intuitive Surgical, both robotics companies in the healthcare industry, and stating that others suggested such
companies were asserting patents to broad to be valid in order to protect their market share.
68 Thayer, L., and Bhattacharyya, A., Patent Eligibility of Software in the Wake of the Alice Corp. v. CLS Bank
Decision, Robotics Business Review, August 14, 2014, and Thayer, L., and Bhattacharyya, A., Will Supreme
Court Rein in Software Patents? Robotics Business Review, March 4, 2014.
On the other hand, however, innovative firms invest in their own R&D efforts and look to
protect their inventions far more vigorously when it comes to those elements of robotics
innovation that differentiate end-products.
This parallel application of cooperative and competitive approaches results in a coexistence
of competitive and open source-inspired approaches to handling IP.
Various non-profit organizations and projects support the development, distribution and
adoption of open-source software for use in robotics research, education and product
development. The iCub, for instance, is an open-source cognitive humanoid robotics
platform funded by the EU which has been adopted by a significant number of laboratories.
Poppy is an open-source platform developed by INRIA Bordeaux for the creation, use and
sharing of interactive 3D-printed robots. Other examples include the Dronecode project and
the NASA International Space Apps Challenge.
Some of this will entail an increasing shift toward engaging end-users or amateur scientists
to interact and improve on existing robotics applications. In fact, many user-oriented low-
cost platforms built for home or classroom use, like TurtleBot and LEGO Mindstorms, are
built on open-source platforms.
This open-platform approach is not limited to software; it can also encompass blueprints
such as technical drawings and schematics, including designs. The Robotic Open Platform
(ROP), for instance, aims to make hardware designs of robots available to the robotic
community under an Open Hardware license; advances are shared within the community.
In general, it will be interesting to see how well the robotics innovation system can preserve
its current fluid combination of proprietary approaches for those aspects of IP where the
commercial stakes are higher plus non-proprietary approaches to promote more general
aspects of relevant science through contests but also collaboration among young roboticists
and amateurs interested in open-source applications.69
4.3 Protecting robotic breakthroughs via technological complexity and secrecy
Potentially more important than patents, the technological complexity and secrecy of robotics
systems are often used as a key tool to appropriate innovation. This is true for standard
mechanical, hardware-related components.
There are multiple reasons why a robotics company may prefer to keep certain technologies
or information as trade secrets rather than seeking patent protection. The first two reasons,
which are not necessarily unique to robotics, are based on either the inability to obtain patent
protection or it is believed that the IP cannot be reverse engineered by even the most
sophisticated competitors. This is frequently done with robotics companies who believe that
their manufacturing process could not be identified without a competitor actually observing
the manufacturing process. The same is frequently true for the methods of testing a robot’s
performance. Some robotics companies have also survived with relatively few competitors
and have the opinion that their work, although comprised of software and hardware, is so
advanced that only a select few could reverse engineer their products.
69 In light of the high degree of collaboration in the robotics industry, as well as many nations' interest in fostering
robotics innovation, it is timely for nations to re-examine their laws on joint IP ownership. Many nations' laws on
joint IP ownership produce unexpected and unfair results in practice, unless the persons or entities involved
contract around such laws.
Robotics companies that make a limited number of highly expensive robots, including for
military applications, typically do not fear that competitors will gain physical possession of
such robots to reverse engineer them. As such, reverse engineering may be difficult for that
practical reason as well.70
Additionally, many small and mid-size robotics companies prefer trade secrets because they
want to avoid the costs and fees that come with filing patent applications.
There are also historical reasons why robotics companies choose to retain information as
trade secrets. In the 1980s, robotics made several significant advances and firms filed a
large number of patents (see figures 3 and 5). However, few of these inventions were
commercialized quickly. As a result, firms spent large amounts of money to obtain patents
that expired before their products were commercialized. They learned from this experience
that patents can be costly without necessarily bringing any reward, especially for innovations
that may be decades away from use in a market-ready product.
Trade secret protection is also important when employee mobility is high. Robotics
companies have sought to protect their trade secrets when an employee joins a
competitor.71 Many robotics companies bolster the protection of their trade secrets with
restrictive covenants to the extent the relevant jurisdiction permits.72
Finally, the more recent questions around the patentability of software in the US and
elsewhere could increase the incentive to protect related inventions via secrecy instead.
4.4 The role of being first-to-market, reputation and strong brands
Being first to market, a strong after-sales service, reputation and brand have all been critical
in past robotics innovation, and they remain so today – all the more so as the industry moves
out of factories and into applications with direct consumer contact.
In the case of industrial automation, only a few trusted operators able to produce a large
number of reliable robots and to service them dependably were in demand by automotive
companies. Initially, Unimation dominated the supply of industrial robots; later, large firms
such as Fanuc held sway.
70 See Keisner (2013b).
71 See e.g. supra, MAKO Surgical Corp.’s lawsuit against Blue Belt Technologies and an employee who left
MAKO Surgical for a position at Blue Belt; see also, e.g., supra ISR Group’s lawsuit against Manhattan Partners.
72 See Keisner (2013a) discussing, among other topics, trade secrets and the use of non-compete agreements
within the robotics industry. One such example is the 2013 lawsuit ISR Group v. Manhattan Partners. Based on
the documents filed in the lawsuit, it appears that ISR Group and Manhattan Partners were in serious talks about
a potential acquisition, but after the deal fell apart and two ISR employees left to join Manhattan Partners, ISR
sued for trade secret misappropriation in Tennessee state court. Manhattan Partners quickly removed the case
to federal court, and then the parties settled. This shows the important trend of robotics companies’ increasing
willingness to defend their trade secrets through litigation when necessary. Another example is the 2013 lawsuit
MAKO Surgical v. Blue Belt Technologies. MAKO Surgical filed a lawsuit in March of 2013 against Blue Belt
Technologies and Blue Belt’s then-recently hired employee for trade secret misappropriation and for the
employee’s violation of his contractual non-compete obligations to his former employer, MAKO. See The
Sincerest Form of Flattery…, Robotics Business Review, March 21, 2013 (discussing the allegations in MAKO
Surgical’s trade secret misappropriation and breach of contract complaint against Blue Belt Technologies).
MAKO obtained a preliminary judgment against Blue Belt Technologies and its employee (See MAKO Wins
Permanent Injunction Against Blue Belt Technologies, Robotics Business Review, April 9, 2013), which
prevented that employee from working for Blue Belt Technologies for a certain period of time. See Keisner
(2013a). However, it should also be noted that, after MAKO Surgical’ s lawsuit against Blue Belt Technologies
for trade secret misappropriation was settled, MAKO Surgical again sued Blue Belt Technologies in June 2014
for patent infringement. See Cole E., “MAKO Surgical Sues Blue Belt Technologies”, Robotics Business Review,
June 10, 2014.
While the landscape is more diverse today, being first and having a solid reputation and
brand continue to be critical. Actors such as hospitals, educational institutions and the
military will want to rely on experienced robotics firms and trusted brands. In the area of
medical robot makers, examples are the DaVinci surgical robot, the CorPath vascular
surgery robots and the Accuray CyberKnife Robotic Radiosurgery System. Even in fields
related to military or similar applications, brands matter, as evidenced by the use of
trademarks such as Boston Dynamics’ “BigDog”. But strong brands are particularly
important when robots are sold directly to end-users; for example, the “Roomba vacuum
cleaner” relies strongly on its trademark value.
Most robotics companies trademark their company names and robot names, with the result
that a growing number of trademarks include the term “robot”.
4.5 Industrial design rights, trade dress, and robotics
Next to patents, industrial designs protecting the ornamental features of a robot – as
registered IP forms – also play an important role in helping firms appropriate the returns to
their investments in R&D.
Another form of IP is trade dress, a source-designating type of IP that refers generally to the
total image of a product.73 Within the robotics industry, trade dress is a right generally used
to describe the total image of a robot or robotics product.74 Although some nations do not
distinguish between trade dress rights and trademark rights, as both are source-identifying
forms of IP, some nations provide protection for trademark rights under a sufficiently broad
definition that it is perceived as extending to other source-identifying forms of IP, including
registered trade dress rights.75
Within the robotics industry, there are only a few examples of lawsuits asserting trade dress
infringement claims based on the “look and feel” of a robot.76 However, there are no known
cases in which such a trade dress infringement claim has been decided on the merits.77
4.6 Copyright and robotics
The Copyright protection is relevant to robotics too, in several respects.
The best example of where robotics companies seek copyright protection is for software
code that has been “reduced to writing” and is believed to be unique and original.78 In
73 See Reese (1994); See also United States Patent and Trademark Offices’ Trademark Manual of Examining
Procedure (2014), Section 1202.02, entitled “Registration of Trade Dress.”.
74 Similar to trade dress rights, industrial design rights protect the visual design of an object. As a result of The
Hague Agreement Concerning the International Deposit of Industrial Designs, there is now a procedure for
international registration of an industrial design, effective in several countries, via a single application. See WIPO
75 See India’s Trade Marks Act, 1999; see also Tiwari (2005), pp. 480 discussing that Indian courts have shown
the propensity to address the issues of trade dress protection within the board parameters of the law on passing
76 See e.g., iRobot Corporation v. Urus Industrial Corporation, et al., 1:05-cv-10914 (D. Mass. 2005) (lawsuit filed
in the United States District Court for the District of Massachusetts by iRobot against Koolatron and Urus Indus.
for alleged trade dress infringement, copyright infringement, and patent infringement); See also, Innovation First,
Inc., et al. v. Urban Trend, LLC, et al., 3:10-cv-934 (N.D. Tx. 2010) (lawsuit filed in the United States District
Court for the Northern District of Texas by Innovation First against Urban Trend for alleged trade dress
infringement, false designation of origin, false advertising, copyright infringement).
77 See e.g., iRobot Corporation v. Urus Industrial Corporation, et al., 1:05-cv-10914 (D. Mass. 2005) (lawsuit
settled and consent judgment filed); See also, Innovation First, Inc., et al. v. Urban Trend, LLC, et al., 3:10-cv-934
(N.D. Tx. 2010) (lawsuit settled).
practice, robotics companies typically use copyright enforcement to prevent others from
copying, or simply accessing, their computer code.
It is generally accepted within most nations that circumventing an electronic barrier in order
to gain access to copyrightable computer code is a violation of the 1996 WIPO Copyright
Treaty.79 This is particularly important to the robotics industry because most robotics
companies employ electronic barriers to restrict access to their robot’s computer code. In
the United States, the case law over the past several years also shows a trend suggesting
that the United States, which is a signatory to the WIPO Copyright Treaty, will conform to the
laws of most other nations and Article 11 of the WIPO Copyright Treaty, such that
circumventing an electronic barrier in order to access copyrightable code – even if there is
no act of copying giving rise to an independent claim for copyright infringement – would still
constitute a violation of the United States’ Digital Millennium Copyright Act (DMCA) 80, which
is the United States’ implementation of the WIPO Copyright Treaty.81 The European Union
has similarly taken measures to harmonize its laws with the WIPO Copyright Treaty by
prohibiting the circumvention of electronic barriers in order to access protected copyrightable
works. The EU has taken such steps by implementing Article 6 of European Directive
Aside from disputes among companies, and despite the fact that some nations laws may
provide for reverse engineering exceptions, copyright related anti-circumvention laws have
also been invoked when an amateur scientist decrypts and changes software code.
Although it was never the subject of a court decision, a company’s assertion of a US Digital
Millennium Copyright Act (DMCA) violation for the unauthorized accessing of the company’s
robot’s software code is not new. When a consumer hobbyist decrypted the software code
for Sony’s robotic-dog, Aibo, and circulated the new software to other consumers such that
they could “teach” the robotic-dog to dance and speak, among other things, the company
asserted that such acts were a violation of the DMCA and demanded the removal of the
4.7 What happens to inventions or creative works produced by robots, now and
A question that cannot yet be considered settled law in any nation, but for which IP
practitioners around the globe may soon face, is whether IP can be created by a robot, and if
so, who owns IP created by a robot?84
78 Another example of where copyright protection could be used for robotics, but is not necessarily a common
practice in the industry is for a unique aesthetic design, such as a design pattern on a robot.
79 See WIPO Copyright Treaty, Article 11 (1996) (prohibiting the circumvention of technological measures for the
protection of copyrightable works).
80 See Keisner (2012) citing MDY Indus. LLC v. Blizzard Entertainment Inc., 629 F.3d 928 (9th Cir. 2010).
81 See U.S. Senate Committee on the Judiciary, Report to the Senate on the Digital Millennium Copyright Act of
1998 (105 S. R.t. 105-190) (“Title I will implement the new World Intellectual Property Organization (WIPO)
Copyright Treaty and the WIPO Performances and Phonograms Treaty, thereby bringing U.S. copyright law
squarely into the digital age and setting a marker for other nations who must also implement these treaties”).
82 See Directive 2001/29/EC of the European Parliament and of the Council of 22 May 2001, Official Journal of
the European Communities, L 167/10, 22.6.2001.
83 See Mulligan & Perzanowski (2007) citing to multiple news sources and discussing Sony’s assertion of a
DMCA violation in stopping the use of unauthorized computer code distributed for use with its Aibo robotic-toy
84 Such “open” issues about robot-generated works and IP rights were the subject of Christophe Leroux’s
presentation at the November 2012 International Workshop on Autonomics and Legal Implications, see Leroux
In the future, robots set to accomplish a task are likely to produce new solutions to problems
and in so doing create physical or intangible products or outputs that could, at least in
theory, be perceived as IP – new inventions, creative works or trademarks, for instance.
This element of robotics innovation could raise interesting questions as to the set-up and
boundaries of the current IP system. Are objects, software code or other assets created
autonomously by a robot copyrightable or patentable? If so, how? And who would own
these IP rights? The producer? The user of the robot? The robot itself?85 Some countries
such as Japan and the Republic of Korea are actually considering extending rights to
Some case law is relevant too: For instance, it was recently determined by the US Copyright
Office that a photographer did not own the copyright to a photograph that was taken by a
monkey who temporarily “borrowed” his camera.86 Given the ruling, some practitioners
question whether photographs taken by robots would be protected by copyright.87 Based on
the language in the U.S. Copyright Office’s recent Compendium, it appears that there would
be no copyright protection in the United States for a work created by a robot.88
In the UK, on the other hand, there is dedicated legislation suggesting that copyright
protection will be avoidable for robot-generated works.89 Although it is debated how such
legislation should be applied,90 it is nonetheless an area of IP law applicable to robotics in
which it appears that contradictory rules are emerging between nations with significant roles
in the robotics industry.
In New Zealand, the law suggests that original works otherwise protectable if created by a
human, are still eligible for copyright protection under New Zealand’s Copyright Act of 1994
even if the work is instead created by software, robots or artificial intelligent systems.91
Ownership of such works, however, would belong not to the robot or intelligent system, but
instead to the person(s) who created or utilized the robot or intelligent system that ultimately
created the work. In comparison, however, IP practitioners in Australia have noted, in light
of some of the same case law referenced by New Zealand practitioners that the laws, while
providing for copyright protection of computer generated works, still involve numerous
aspects that make it practically difficult to assert copyright protection over a work generated
by a computer, robot or intelligent system.92
A full legal assessment of this question relating to autonomous robot creation is beyond the
scope of this report, but who owns the IP rights over creations produced by robots will surely
be a matter of much future discussion.
86 See Compendium of U.S. Copyright Office Practices, Third Edition (Public Draft – Not Final), August 19, 2014
clarifying that photographs taken by a monkey are not eligible for copyright protection; See also David McAfee,
Copyright Office Says It Will Not Register ‘Monkey Selfie’, Law 360, August 22, 2014.
87 See Mark A. Fischer, Are Copyrighted Works Only by and for Humans? The Copyright Planet of the Apes and
Robots, Duane Morris New Media and Entertainment Law Blog, August 18, 2014.
89 See RoboLaw, Guidelines Regulating Robotics, p. 19 (2014).
91 See Simpson Grierson, Earl Gray, Raymond Scott, If Shakespeare were a robot, would “he” be an “author”?,
lexology.com, February 28, 2011 discussing New Zealand’s Copyright Act and recent decisions by the Australian
Federal Court suggesting that original works are copyright protected in New Zealand.
92 See Tim Clark, Ivor Kovacic, Copyright in works generated by computer programs, Lexology.com, August 31,
Few studies have analyzed the developments of robotics, the underlying innovation eco-
system and the role of IP. This paper fills this gap by providing an up-to-date analysis of the
robotics innovation system. Original patent landscape data are presented to shed light on
robotics filing strategies and to identify top filers. The paper goes beyond the use of patents
to also cover the role of trade secrets, industrial design, brands and copyright. To conclude
the paper also shows how the emergence of robots could lead to new questions such as
who owns the IP of works or inventions created by robots themselves?
Brynjolfsson, E., and McAfee, A. (2014). the Second Machine Age: Work Progress, and
Prosperity in a Time of Brilliant Technologies. New York:: W. Norton and Company.
Christensen, Henrik, Goldberg, L. Kumar, V. and E. Messina (2013), A Roadmap for U.S.
Robotics—From Internet to Robotics, Robotics-Virtual Organization, Report, workshop
“Manufacturing and automation robotics”, Washington D.C., October 2,
Cooper, D. M. (2013). A Licensing Approach to Regulation of Open Robotics, Paper for
presentation at We Robot: Getting Down to Business Conference, Stanford Law School,
April 8-9, 2013.
euRobotics (2014). Strategic research agenda for robotics in Europe 2014-2020. Retrieved
August 10, 2015, from http://www.eu-robotics.net/cms/upload/PPP/SRA2020_SPARC.pdf.
Frey, C. B., and Osborne, M. A. (2013). The future of employment: How susceptible are jobs
to computerisation? (Working Paper Oxford Martin Programme on the Impacts of Future
Technology). Oxford: Oxford University.
IFR. Service Robots. Retrieved August 3, 2015, from International Federation of Robotics:
IFR. (2012). History of Industrial Robots: From the first installation until today. Frankfurt am
Main: International Federation of Robotics.
IFR. (2014a). World Robotics 2014 Industrial Robots. Frankfurt am Main: International
Federation of Robotics.
IFR. (2014b). World Robotics 2014 Service Robots. Frankfurt am Main: International
Federation of Robotics.
Keisner, C. A. (2012), Making Your Robotics Company A More Attractive Investment,
Robotics Business Review, October 21, 2012.
Keisner, C. A. (2013a), Robotics and Intellectual Property, Incremental Advantage Webinar,
Keisner, C. A. (2013b), Keeping Things Confidential: Robotics Trade Secrets 1.0. Robotics
Business Review, October 21, 2013.
Keisner, C. A. (2015), Report for WIPO on intellectual property and robotics. Unpublished
background report for the World Intellectual Property Report 2015.
Kumaresan, N., & Miyazaki, K. (1999). An integrated network approach to systems of
innovation—the case of robotics in Japan. Research Policy, 28(6), 563-585.
Leroux, C. (2012). EU Robotics Coordination Action: A green paper on legal issues in
robotics. Paper presented at the International Workshop on Autonomics and Legal
McGurk, M. R., & Mandy, J. (2014). Building a Strong Robotics IP Portfolio.
McKinsey Global Institute. (2013). Disruptive Technologies: Advances that will Transform
Life, Business, and the Global Economy.
Metra Martech. (2011). Positive Impact of Industrial Robots on Employment. London: Metra
Miller, B., & Atkinson, R. D. (2013). Are Robots Taking Our Jobs, or Making Them?
Washington, D.C.: The Information Technology and Innovation Foundation.
Mireles, M. S. (2006). States as Innovation System Laboratories: California, Patents, and
Stem Cell Technology. Cardozo Law Review 1133(374).
Mulligan, D. K., and A.K. Perzanowski (2007 ). The Magnificence of the Disaster:
Reconstructing the Sony BMG Rootkit Incident. Berkeley Technology Law Journal 22(1157),
Nobile, C. (2013), The IP Battle Continues For Robotics Companies, Robotics Business
Review, January 7, 2013.
Nof, S. Y. (1999). Handbook of Industrial Robotics. West Sussex Wiley.
RBR Staff (2013). iRobot Obtains Injunction against Shenzhen Silver Star, Robotics
Business Review, September 6, 2013.
Reese, J. W. (1994), Defining the Elements of Trade Dress Infringement under Section 43(a)
of the Lanham Act, 2 Tex. Intell. Prop. L.J. 103 (1994).
Rosheim, M. E. (1994). Robot Evolution: The Development of Anthrobotics: John Wiley &
Scheinman, V. (Producer). (August 10, 2015) Robotics History Narratives. Interview
retrieved from http://roboticshistory.indiana.edu/content/vic-scheinman.
Siegwart, R. (2015). Report for WIPO on intellectual property and robotics. Unpublished
background report for the World Intellectual Property Report 2015.
Smith, R. C., & Cheeseman, P. (1986). On the Representation and Estimation of Spatial
Uncertainty. The International Journal of Robotics Research, 5(4), 56-68.
Springer, P. J. (2013). Military Robots and Drones: A Reference Handbook. Santa Barbara:
Technopolis and University of Manchester (2011), Case Study on the demand-side elements
of the Danish innovation policy
Tiwari, A. (2005). Passing off and the Law on ‘Trade Dress’ Protection: Reflections on
Colgate v. Anchor, Journal of Intellectual Property Rights, West Bengal National University
of Juridical Sciences, Vol. 10, pp. 480-490.
Tobe, F. The Robot Report's Global Map. Retrieved August 10, 2015:
UKIPO. (2014). Eight Great Technologies - Robotics and Autonomous Systems: A Patent
Overview. London: UK Intellectual Property Office.
Wintergreen Research Inc. (2015). Surgical Robots Market Shares, Strategies, and
Forecasts, Worldwide, 2015 to 2021. Dublin: Research and Markets.
World Intellectual Property Organization (WIPO) (2014). Guide to the International
Registration of Industrial Designs under the Hague Agreement, WIPO Publication No.
857(E), ISBN 978-92-805-2522-9 (WIPO 2014 2nd edition), Geneva: WIPO.
WIPO (2011), Harnessing Public Research for Innovation – the Role of Intellectual Property,
Chapter 4, in World Intellectual Property Report 2011 - The Changing Face of Innovation,
WIPO (2015), World Intellectual Property Report 2015 - Breakthrough Innovation and
Economic Growth, Geneva: WIPO.
Part of the empirical analysis of this paper relies on a tailor made patent mapping. The
patent data for these mappings come from the WIPO Statistics Database and the EPO
Worldwide Patent Statistical Database (PATSTAT, April 2015).
The patent mapping strategy was adapted from the seminal work by the UKIPO93 combining
CPC and IPC symbols with text terms searched for in titles and abstracts. In particular, we
have included the following list of IPC and CPC symbols: B25J 9/16, B25J 9/20, B25J
9/0003, B25J 11/0005, B25J 11/0015, B60W 30, B60W2030, Y10S 901, G05D 1/0088,
G05D 1/02, G05D 1/03, G05D2201/0207, and G05D 2201/0212; complemented with the
following terms: robot, robotics, and robotic.
The resulting sample was benchmarked against a list of seminal patents and a list of
robotics companies. The latter was compiled – along with their geographic locations and the
company type –based on information about robotics companies available to robotics-focused
associations and groups, including The Robot Report’s Global Map, as well as from the
publicly available listing of companies from the Robotics Industry Association (RIA). While
these sources are useful to corroborate the location of robotics companies, and the
formation of robotics clusters, which supports the perceived importance of collaboration
within the industry, the identification of robotics companies by such robotics-focused
associations and groups has certain shortcomings in the way that they are used herein.
Nonetheless, it appears that the shortcomings are minor and do not have a significant
impact on the conclusions derived from the data (see Figure 2 and footnote 29).
The main unit of analysis is the first filing of a given invention.94 In consequence, the date of
reference for patent counts is the date of first filing. The only departure from this approach
occurs when analyzing the share of patent families requesting protection in each patent
office in figure Figure 7. In this case, an extended patent family definition – known as the
INPADOC patent family – has been used instead of the one relying on first filings. In
addition, only patent families with at least one granted application have been considered for
this analysis, and the date of reference is the earliest filing within the same extended family.
The main rationale for using the extended patent family definition and imposing at least one
granted patent within the family is to mitigate any underestimation issuing from complex
subsequent filing structures, such as continuations and divisionals, and from small patent
families of lower quality such as those filed in only one country and either rejected or
withdrawn before examination.
The origin of the invention is attributed to the first applicant of the first filing; whenever this
information was missing an imputation strategy has been applied. When information about
the first applicant’s country of residence in the first filing was missing, the following sequence
was adopted: (i) extract country information from the applicant’s address; (ii) extract country
information from the applicant’s name (see further below); (iii) make use of the information
from matched corporations (as described further below); (iv) rely on the most frequent first
applicant’s country of residence within the same patent family (using the extended patent
family definition); (v) rely on the most frequent first inventor’s country of residence within the
same patent family (again, using the extended patent family definition); and (vi) for some
remaining historical records, consider the IP office of first filing as a proxy for origin.
Applicants have been categorized in three broad categories: (a) Companies, which includes
mostly private companies and corporations, but also state-owned companies; (b) Academia
93 See UKIPO (2014).
94 Mappings include data on utility models whenever available.
and public sector, which includes public and private universities (and their trustees and
board of regents), public research organizations, and other government institutions such as
ministries, state departments and related entities; (c) Individuals, which includes individual
first applicants who may or not be affiliated with companies, academia or other entities. A
further category, (d) Not available, includes all unclassified first applicants.
In order to assign broad type categories to each first applicant, a series of automated steps
were performed for each of the six innovation fields underlying the case studies, to clean and
harmonize applicant names. The results of this automated process were cross-checked
manually – particularly for the top applicants of each type – prompting revision of the
strategy and adjustment of parameters in several iterations.
The starting point was the original information about the first applicant’s name from the first
filing. When this name was missing, the most frequent first applicant’s name within the same
patent family using the extended definition was considered. This list of improved first
applicants’ names was automatically parsed in several iterations in order to: (i) harmonize
case; (ii) remove symbols and other redundant information (such as stop words and
acronyms); (iii) remove geographical references (used to improve information on applicants’
country of residence); and (iv) obtain any valuable information on applicant names meeting
criteria to be considered as (a) companies or (b) academia and public sector types.
Subsequently, a fuzzy string search was performed – using Stata’s matchit command95 – in
order to detect alternative spellings and misspellings in applicant names, and the types were
propagated accordingly. In addition, the results of corporation consolidation (see below) also
permitted recovery of some unclassified applicant names as companies. Finally, the
category individuals was imputed only to remaining unclassified records when they either
appeared as inventors in the same patent or were flagged as individuals in the WIPO
Statistics Database for patent families containing a PCT application. Analysis of the
unclassified records indicates that most of them have missing applicant names in PATSTAT.
Most of these missing names refer to original patent documents not in Latin characters and
without subsequent patent filings.
The rankings provided for robotics consolidate the patent filings of different first applicants.
Manual examination and consolidation was performed for the most frequent applicants.
Applicants sharing a common ultimate owner were consolidated into one. In the case of the
top 30 companies, the ownership profiles in the BvD Ownership Database were used. Only
subsidiaries that were directly or indirectly majority owned were taken into account in the
95 Available at the Statistical Software Components (SSC) archive and from the WIPO website.