Variety of technological trajectories in low emission vehicles (LEVs): A patent data analysis

Journal of Cleaner Production 17 (2009) 201–213

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

Variety of technological trajectories in low emission vehicles (LEVs): A patent data analysisq Vanessa Oltra, Maı¨der Saint Jean* GREThA (UMR CNRS 5113), Bordeaux IV University, Avenue Le´on Duguit, 33608 Pessac, France

a r t i c l e i n f o

a b s t r a c t

Article history: Received 22 February 2007 Received in revised form 22 April 2008 Accepted 24 April 2008 Available online 17 June 2008

This paper focuses on the diversity of engine technologies for low emission vehicles (LEVs) that are developed by car manufacturers in order to substitute for the conventional internal combustion engine vehicle (ICEV). Our purpose is to analyse the competition between the various technologies for LEVs as well as the innovative strategy of car manufacturers. We first propose to define and to represent these technological trajectories in order to compare their performances and to identify their strengths and weaknesses. The technological bottlenecks, the barriers to the adoption of these alternative engine technologies as well as the features of this technological competition are underlined. We then use a patent data analysis to study the patent portfolios of the main car manufacturers in these technologies over the period [1990–2005]. The dynamics of patent applications by car manufacturers gives insight on the competition among technologies and on the strategy of firms. This analysis emphasises the progressive diversification of firms’ patent portfolios over the whole set of engine technologies and the differentiated strategic positioning of car manufacturers according to countries. Ó 2008 Elsevier Ltd. All rights reserved.

JEL: O33 Q55 L62 Keywords: Environmental innovation Technological competition Diversity Low emission vehicles Patents

1. Introduction Under the pressure of regulation, the automotive industry has to cope with environmental concerns, in particular with the reduction of polluting emissions (carbon monoxide, nitrogen oxide, particle matters, etc.), fuel consumption and noise, as well as with the recycling of end of life vehicles. This regulatory context has encouraged R&D and innovative activities of car manufacturers and their suppliers. The Zero Emission Vehicle (ZEV) Mandate introduced by the Californian Air Resources Board in 1990 gave an important impulse to the development of low emission vehicles (LEVs). This technology-forcing regulation primarily focused on electric vehicles. But electric vehicles did not lead to a sizeable market and their commercialisation failed due to the unsatisfying performance characteristics of battery technology compared to mainstream technologies. That is the reason why other technologies started to be supported, such as fuel cell vehicles and hybrid

q Support for this research by the CCRRDT programme of Aquitaine Region is gratefully acknowledged. * Corresponding author. E-mail addresses: [email protected] (V. Oltra), [email protected] (M. Saint Jean). 0959-6526/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.jclepro.2008.04.023

vehicles. This evolution leads to a technological competition between the different technologies for LEVs.1 In this paper, we focus on the diversity of engine technologies for LEVs that are developed by car manufacturers in order to substitute for the conventional internal combustion engine vehicle. Our purpose is to analyse the competition among these LEVs technologies as well as the innovative strategy of car manufacturers. In the first section, we propose to define and to represent these technological trajectories in order to compare their performances and to identify their strengths and weaknesses. We discuss the technological bottlenecks, the barriers to the adoption of these alternative engine technologies, as well as the features of this technological competition. In Section 2, we use a patent data analysis to study the patent portfolios of the main car manufacturers in these technologies over the period [1990–2005]. The dynamics of patent applications gives insight on the competition among technologies and on the strategy of firms. This analysis emphasises the progressive diversification of firms’ patent

1 In the line of Brian Arthur’s works ([1,2]), technological competition refers to a situation where rival variants of the same technology are being developed by competing firms while their adoption benefit from increasing returns to adoption may lead to a monopoly of one single technology and thus to a technological lock-in.

202

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

portfolios over the whole set of engine technologies and the differentiated strategic positioning of car manufacturers according to countries. 2. Technological variety and competition among LEVs 2.1. The different competing technologies for LEVs Under the pressure of environmental regulation, in particular the Zero Emission Vehicle Mandate introduced by the Californian Air Resources Board in 1990, car manufacturers try to develop more efficient engine technologies (see for example Ref. [3]). The aim is to develop low emission vehicles (LEVs) which are able, in the medium and long term, to comply with environmental standards. The main environmental concerns are greenhouses gases (especially CO2) and local emissions (NOx, CO, VOC), fuel consumption, energy efficiency and engine noise. These environmental concerns have created new technological opportunities in the field of motor vehicles and engines, which lead to an intense activity of environmental innovations. As a result, different engine technologies are explored and developed in parallel by car manufacturers. Our purpose is to analyse how these various technologies compete and how car manufacturers try to combine the different technological options. We first present and discuss these various engine technologies. 2.1.1. Internal combustion engine vehicles (ICEVs) The car industry is characterised by a strong and persistent dominant design which is the internal combustion engine vehicle (ICEV). Since ICEV is already a very mature technology, it is unclear whether this technology is able to meet future environmental regulations and emission standards. Nevertheless during last decades, ICEVs have been significantly improved by innovations like direct injection technologies, the Common Rail technology, Stop and Go systems, particle filters and new materials to lighten vehicles and to decrease frictions. So a cluster of innovations has been developed which has enabled to decrease fuel consumption, polluting emissions and noise rate, and to increase energy efficiency of engines. Gasoline engines have benefited from continuous improvements in order to optimise their environmental performance. Advances have been made in controlling the activity occurring within the engine combustion chambers, through the development of technologies for ignition control and improved air–fuel mixing. Newer possibilities include direct petrol/diesel injection, various valve inlet geometry devices and variable compression systems, all aiming at redefining dynamically the engine power curve to match current conditions. The adoption of these technologies would see the emissions’ performance of fairly standard petrol-powered vehicles approach that of LEVs [4]. This is also true for diesel engine vehicles which environmental performance has strongly been improved. In Europe, diesel vehicles represent an important share of the market in some countries: the share of diesel cars in first registration of passenger cars in 2002 is 63% in France, 57% in Spain, 43% in Italy, 23.5% in UK, 15% in Finland and 0.7% in Sweden [5]. These figures reveal the differences among countries resulting from different technological choices. In contrast, diesel cars are less present in the American market because the incentive to buy is much weaker (due to a smaller price premium for diesel in comparison with gasoline) and due to the fact that diesel cars have difficulties to meet the environmental standards. The European and Japanese car manufacturers are leading in the production of diesel technologies and related innovations.2 Resistance in the US and in Japan has been motivated by the

2 Japanese manufacturers produce diesel cars mainly for the European market and the ‘third world’ market.

potentially carcinogen properties of microscopic soot particles in diesel exhaust gases, by their high noise levels and comparatively poor acceleration, although recent innovations have largely addressed both these issues [6]. Now diesel cars are far more powerful, refined and quiet, and so offer a particularly attractive solution for reducing fuel consumption and carbon dioxide emissions. The future development of diesel cars will depend on fuel availability and fuel price, but also on their environmental performances. We can even find signs of a revival of interest in the US for diesel engines which leads to forecast an increase in diesel penetration in the future 10 years [6]. In Europe, the ERTRAC strategic research agenda also emphasises that ‘‘In the time period to 2020, the main improvements in energy use and GHG emissions will come from efficient ICE and their associated advanced fuels’’ ([7], p. 42). The main research topics planned in this agenda are advanced ICE (high specific torque, advanced fuel injection, flexible components,.), new combustion concepts (controlled autoignition, extended homogeneous range in diesel engines, integrated combustion process,.), fuels and lubricants for advanced ICEs, improved design elements and biomass derived fuels. Thus, the dominant design of ICEV is still under progress even if the technology is considered to be mature. In terms of environmental performance, this technology competes with three alternative engine technologies that are electric vehicles, fuel cell vehicles and hybrid vehicles. 2.1.2. Electric vehicles (EVs) In the beginning, the Californian ZEV mandate created a window of opportunity for EVs but did not lead to a sizeable market for such vehicles [3]. The advantages of electric engines compared to conventional ones are that they do not emit any emissions during use, quiet3 and have less moving parts which reduces the need for maintenance. But the main disadvantage is that the energy supply required to power the vehicle needs to be stored as electricity on board. The batteries that are needed are expensive, have a short lifecycle and have a limited storage capacity. The range of use of these vehicles, i.e. their radius of operation for long distance trips, is therefore much smaller than for other conventional ones, and in particular they are not adapted to intensive use. A large amount of research has been done on alternative systems of battery but with no sign of a breakthrough. Their much greater cost is not sufficiently compensated by gains in power density, battery life or speed of recharging. That is the reason why EVs are not considered to be a direct competitive alternative to ICEV for some applications involving long distance trips. Consequently, the regulatory requirements have been adapted and the window of opportunity has changed to hybrid and fuel cell vehicles. Nowadays, battery electric vehicles remain an area of research and production that involves some new collaborative partnerships between the automotive industry and new partners coming from the electric and electronics industry. However, EVs are limited to niche markets dedicated to specific uses such as delivery vehicles, airport shuttles and urban buses. Thus, EVs first considered in the beginning of the 1990s as an alternative to conventional vehicles, which can be mandated by law with the ZEV mandate, became an alternative technology restricted to certain market niches and captive fleets. R&D activities and patents of automotive firms confirm this evolution since we can observe a rise and fall of EVs patents between 1990 and 2002 with a top in 19964 [8,9].

3 Although electric drivetrains are quiet compared with internal combustion engines, other sources of noise generated by vehicles are important to consider. In particular, the road traffic noise generated by air moving over the vehicle, and especially by tyres in contact with the road, can make difficult to substantially reduce the noise limit imposed on cars. 4 The results presented in Section 2 shows the same trend of evolution of EVs patents.

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

2.1.3. Fuel cell vehicles (FCVs) A FCV is defined as a vehicle driven by an electric engine which is powered by a fuel cell. A fuel cell is an electrochemical cell which can convert the chemical energy of a fuel such as hydrogen into electrical energy. In terms of environmental performance, when the fuel is hydrogen the system does not generate any pollutants. The only emissions are unused oxygen, nitrogen and water, which may be present in either liquid or vapour form. FCVs also have the same advantages as a typical electric engine that is the possibility to regenerate ‘breaking energy’, the omission of transmission and the low noise rate. One major issue with this technology is the fuelling of vehicles. Hydrogen can be fuelled directly in the vehicle or be produced on board when other fuels like gasoline or methanol are used. In the latter case, fuel conversion leads to unwelcome emissions, involves a complicated technical process and reduces the energy efficiency of the system. Moreover, when FCVs use a fuel converter, the global level of emissions is not significantly lower for hydrogen vehicles than for ICEVs if we account for all the steps from ‘‘well-to-wheels’’ [10]. In fact, a FC engine needs a supply of very pure gaseous hydrogen (impurities inactivate the cell), while a conventional engine simply needs a tank of gasoline or diesel fuel. The economic viability of FCVs is also called into question since they represent the most expensive option for most market segments. Even projections over the next decades show that fuel cell vehicles would cost about 100% more than ICE-PISI vehicles, i.e. ICE vehicles using the Port Injection Spark Ignition technology [10]. This premium in vehicle retail price represents an obstacle to the adoption and diffusion of alternative powertrain technologies. Thus, fuelling infrastructure, costs, reliability and durability are the critical hurdles that FCVs have to overcome before they can achieve their development. The fact that the FC technology needs a new fuel distribution and retailing infrastructure involves huge costs5 and is a big problem for the passenger car fleet which is very widely dispersed. Furthermore, this infrastructure would have to operate in parallel with the conventional fuel distribution network, as it would take decades for the huge fleet of vehicles to completely switch over. Another issue pointed out by Maxton and Wormald [6] is that the great majority of the subsystems and components being developed for fuel cells are based on technologies that are almost wholly alien to the car industry. This will create substantial challenges and the need for carmakers to build up new supplier bases. This will strongly affect the relationship between car manufacturers and their suppliers. All this does not make the widespread deployment of FCVs a wildly attractive proposition, either financially or environmentally speaking. That is the reason why the research activities on FCVs result in prototypes for limited range of use, such as urban transit buses and taxis. As a consequence, it seems that FCVs will remain restricted to captive niche markets for governmental agencies or some specific enterprises. 2.1.4. Hybrid vehicles (HVs) The application of hybrid technology looks very promising. Given present limitations in technologies such as batteries and fuel cells, the most viable powertrain alternatives are hybrid configurations that include a relatively small internal combustion engine and an electric motor. A HV system seeks to operate a conventional engine at maximum efficiency or turn it off, and then provides propulsion through an alternative source. This is more efficient in terms of fuel use and less polluting in terms of emissions. The HV

5 The cost of a full hydrogen distribution network for the US has been estimated as $100 billion [6].

203

can be equipped with either a gasoline fuelled ICE as well as a diesel engine, the latter being more efficient. HVs can have serial or parallel propulsion systems depending on the way the combustion engine, the electric engine and the batteries are connected. These vehicles can be fuelled with conventional or alternative fuels. Compared to a serial system, the parallel system can deliver more power due to the simultaneous use of combustion engine and electric engine. This technology is generally considered to be a transition technology between ICEVs and FCVs. One of the advantages of this technology is that it is competitive, in comparison with the internal combustion engine, in terms of range of use and speed, with a total efficiency which can be twice as high as the efficiency of the combustion engine [11]. Previously positioned as an intermediary solution towards new technologies, the HV has always played a modest role in sustainable mobility. Nevertheless, its recent commercial success and its environmental performance slowly change the perspectives for this technology: instead of being developed side-by-side, HVs might in fact form a competitive threat to the commercial development of FCVs [12]. The main advantage of HVs is that they are compatible both with the available fuel infrastructure and the current ICE system.6 In terms of emissions, HVs score very well, especially when diesel is used as fuel, and a good energy efficiency is realized by harnessing the kinetic energy generated when the car brakes. HVs cover a multitude of possible propulsion architectures, from a pure battery electric with a small ICE powered generator on board to extend its range, to a pure ICE driveline with a bit of additional power and regenerative braking capacity via a special starter alternator. The former is called EV ER (electric vehicle, extended range). Work at the University of California has shown that such vehicles can achieve very low levels of fuel consumption and low levels of emissions [6]. The latter are called electric boost vehicles which approximately use 10% of electricity as an energy source. So HVs correspond to a range of technological options resulting from the hybridisation of a conventional ICE and an electric propulsion system. Each option can be characterised by the share of electricity used as an energy source.7 Thus, HVs also have the advantage of offering a large range of hybrid configurations, thus extending their possibility of use. The flexibility offered by hybrid configurations enables carmakers to seek synergies across the various technologies and opens multiple technology pathways that can be explored. Fig. 1 depicts the evolution of the potential market share of hybrids in EU-25 over the next decades [13]. The projections suggest an S-curve market penetration that could lead to a 6% market share in 2010 and slightly above 12% by 2020.

2.2. Environmental performance and range of use of LEVs In order to be efficient and to further develop, each engine technology has to combine several dimensions in terms of environmental performance, engine efficiency, price and range of use of the vehicle. The ability to combine these dimensions is essential to the development and the diffusion of LEVs. This point is critical for the development of clean technology in general [14]. This argument is even stronger in the case of the automotive industry which produces a complex system product that requires firms to coordinate a broad array of different sources of knowledge and

6 Actually hybrids have been in use for a long time for certain applications (intercity trains, heavy duty off-road trucks used in quarries and mining for example) and since several years, they represent a growing centre of attention as a way to meet the numerous demands (environmental, safety, noise, etc.) placed on an engine used to power a vehicle. 7 For example, the Toyota Prius is around 30% of electric power capacity [6].

204

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

Fig. 1. Potential market shares of hybrids in EU-25 (source: [13]).

technology. Each LEV technology differs according to the performance package (pollution, fuel efficiency, price and range of use) they are able to achieve, but no technology is better on all the criteria. In other words, there is no one best technology. In terms of environmental performance, in particular carbon dioxide emissions, Fig. 2 shows that the hydrogen fuelled FCV and the diesel–hybrid vehicle show the largest potential compared to conventional ICEVs. But the performance of FCVs falls down as soon as one puts an on-board reformer. We can observe that, regarding fuel consumption, the best potential for reduction is obviously FCVs. This figure also shows that advanced diesel vehicles exhibit a very good performance. Finally, the best combination between fuel consumption and carbon dioxide emissions is obtained by diesel–hybrid vehicles. Nevertheless, the development and the diffusion of engine technology strongly depend on the price of the technology and the range of use of the vehicle. The range of use refers to the radius of operation of vehicles, which is more or less extended depending on the urban use for example. As a matter of fact, the range of use is a critical point for LEVs. EVs and FCVs are less suitable for intensive use, thus restricting their development to niche markets. HVs present more flexibility due to the combination of ICEV and electric engines, but they remain more adapted to light vehicles. So it clearly appears that the ability to combine environmental performance with the range of use is a key determinant of the development of LEVs. As represented in Fig. 3, the development of LEVs can be summarised by three technological trajectories characterised by their ability to combine these two dimensions. HVs correspond to what

we called in Oltra and Saint Jean [14] a ‘‘median strategy’’, which is often the most efficient since it associates the environmental performance with other performance characteristics of the product. The performance of the product is often critical in the development of clean technologies since they may entail a destroying competence effect that alters certain characteristics of the product [14]. In the case of the automotive industry, the characteristics that are critically modified by the change in engine technology are the power efficiency of the vehicles and the types of fuel (the problem of fuel infrastructure is tackled in the next section). The resulting range of use of LEVs determines their potential market. The example of EVs illustrates this argument since, in spite of their promising environmental performance, their development has been hindered by the relatively low performance of the vehicle in terms of autonomy and power efficiency. On the contrary, HVs attract a lot of attention and show relative market success, like the Toyota Prius whose sales have considerably increased since 1998 in North America and in Japan. The Toyota Prius II really exhibits a good compromise between power engine, fuel consumption, carbon dioxide emissions, noise rate and drivability of the vehicle, which will certainly boosts its sales. The recent years have shown that the main world leading car manufacturers enter the race to hybrid concept cars taking the opportunity of the international showcases to exhibit their most recent innovations. We can say that for now this hybrid technology is the most promising because of its ability to combine environmental performance with the performance of the vehicle, particularly in terms of range of use. But, even if the HV is a solid competitor to the ICEV, the conventional engine is by no means played out. The diesel ICEV is still under improvement and has significantly increased its environmental performances through innovations such as direct injection and particle filters. European car manufacturers still bet on this technology for the future (in particular Peugeot) since they argue that they will be able with advanced diesel cars to exhibit the same performances than HVs in terms of carbon dioxide emissions and fuel consumption for a very lower price. So the technological competition between these engine technologies is very active and is conducive to different strategic positioning of car manufacturers. 2.3. Technological competition between ICEV and LEVs and among LEVs The development of LEVs corresponds to a typical case of technological competition between an established technology, or

200 Litres per 1000 km

180

CO2 emissions grams/km

160 140 120 100 80 60 40 20 0 Conventional ICE-gasoline

Conventional ICE-diesel

Advanced ICE- Advanced ICEgasoline diesel

Hybrid ICEgasoline

Hybrid ICEdiesel

Fuel cell engineFuel cell compressed engine-on hydrogen board reformer

Fig. 2. Fuel consumption and CO2 emissions for the different engine technologies. Source: [6], p.85.

Environmental performance

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

EVs FCVs

HVs Advanced ICEdiesel

Range of use Fig. 3. Technological trajectories for the development of LEVs.

a dominant design, and a set of alternative technologies. Evolutionary economics emphasises that in that case market selection may select a suboptimal technology as increasing return to adoption renders the process of technology selection path dependent [2,15]. Path dependency can create lock-in on suboptimal technology because self-reinforcement effects stabilise one technology and inhibit the transition towards a new one. Arthur [1] identifies five sources of path dependency which are learning by using, network externalities, scale economies in production, informational increasing returns and technological interrelatedness. As a consequence even if a superior technology is introduced in the market, technological substitution is not warranted because the established technology benefits from increasing returns to adoption. When technological substitution does take place and several new technologies are competing, market selection can select a suboptimal technology due to path dependence of sequential adoption decisions. In the analysis of technological competition, it is often assumed that the competing technologies meet similar functions. This statement is not a trivial one because of the huge amount of features that usually characterise a technology. Whenever a new technology fails to be successful, we therefore have to analyse carefully whether it did so because it was impeded by the dominance of the established technology or because it was simply inferior with regard to the functions it was supposed to fulfil [16]. In other words, technologies often do not serve as perfect substitutes. This notion of functions of technology is helpful to distinguish two kinds of technological competition. In many cases, the solution of an environmental problem defines a new function and several technologies executing this function compete on the level of what Sartorius and Zundel [16] call new-versus-new competition. In order to be accepted by consumers, the new environmentally friendly technologies also have to fulfil the genuine function of the established technology they are supposed to replace. This gives rise to an old-versus-new competition. In our case, this last type of competition concerns the competition between ICEVs and LEVs, while the new-versus-new competition occurs between EVs, FCVs and HVs. Both types of competition need to be analysed in close relation since there are mutual interactions between them. The old-versus-new competition between the ICEV and LEVs can be studied as a problem of competition between a strong dominant design and a set of alternative new technologies. The barriers to the development of new technologies are mainly linked to economies of scale and scope, as well as to learning effects and network externalities. The dependence of the ICEV on a network of filling stations is a source of network externalities. The degree of compatibility of the new engine technologies with the existing network and infrastructures is a crucial factor of this old-versusnew competition. Indeed a significant change in the fuel

205

infrastructure involves huge costs which create strong barriers to the adoption of a radically new engine technology. This is particularly the case for FCVs fuelled with hydrogen. This can lead to strategic behaviours in order to keep the current fuel infrastructure and so to create compatibility between the established technology and the new ones.8 The barriers to the implementation of new engine technologies for LEVs are also linked to the accumulation of knowledge and competencies on the established ICEV technology which creates strong irreversibilities. These irreversibilities are reinforced by the vertical relationships between car manufacturers and their suppliers which can be affected by new engine technologies. In other words, the more radical and the more global the innovation, in the sense that it requires many changes in the linkages between actors, the stronger the implementation barriers [12]. These features tend to favour the persistence of the ICEV dominant design and to spur incremental innovations on this design.9 The new-versus-new competition between EVs, FCVs and HVs is governed by the same forces as the old-versus-new competition. The main difference consists in the point of departure of the competition, since in the former case the initial competitive advantage is fundamentally in favour of the established technology, while the new alternative technologies have to build up progressively their competitive advantage. In the case of LEVs, this new-versus-new competition is really important since various technological options are explored. This exploration of alternative engine technologies leads to a race to innovation and to specific strategic positioning of car manufacturers. The competition between EV, FCV and HC is characterised by an evolution in the priority of research activities and in the strategy of car manufacturers. From 1990 to 1996, the EV was considered to be a serious alternative to ICEV. Given the constraints on batteries and on the range of use of EVs, there was a change in perspectives and FCVs appeared progressively as the most promising technology. Since 1997–1998, the vast majority of the automotive industry has embraced fuel cells leading to large-scale research and development activities [12]. More recently the HV, which was previously positioned as an intermediary solution towards radically new technology such as FCV, developed progressively and plays now a dominant role. The main advantage of the HV is its compatibility with the ICEV dominant design and with the available infrastructure, while the FCV requires changes at both levels. As a result the HV will experience much less barriers than the FCV. This competitive advantage of the HV is linked to the fact that technological hybridising enables to exploit technological complementarities between the internal combustion engine and the electric one. Technological complementarities play an increasing role in this competition among LEV technologies since there exist spillovers between technologies. For example, the improvements of EVs have benefited to HVs, as well as innovations on HVs benefited to the FCV. Moreover we also observe that research activities on fuel cells result in the development of the use of fuel cells for the feeding of electronic and electric devices of ICEVs. As a matter of fact, there is an increasing overlapping of technologies which enables car manufacturers to exploit technological complementarities between alternatives, to progress step by step and to improve continuously the ICEV dominant design. This overlapping of technologies is a way of building a gradual path towards radically new engine technologies. As a result the technological competition among LEVs is characterised by a persistent diversity of options, no signs of

8 An example is the development of a fuel reformer by Shell to convert gasoline in hydrogen on board of the vehicle [12]. 9 Anyway the persistence of ICEV for a rather long time is unavoidable since it will take a very long time to replace the whole car fleet.

206

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

premature lock-in10 and increasing complementarities between technologies which mainly benefit to the HV. 3. An analysis of patent portfolios of car manufacturers 3.1. Method and data Data on patent applications from 1990 to 2005 come from the European network of patent databases (esp@cenet Portal11) based on the European Patent Office (EPO). We used the worldwide database since it covers published patent applications from more than 70 countries and regions and is the most comprehensive collection of documents in esp@cenet. We selected 11 car manufacturers according to the top sales ranking in output units in 2002 [5]. The sample of firms under study thus comprises Toyota, Honda, Renault, Ford, Nissan, Mitsubishi, Hyundai, General Motors, Peugeot, Daimler-Chrysler and Volkswagen. This selected sample is also justified by our purpose to compare the strategies of American, Japanese and European, in particular French and German, car manufacturers. Indeed these three regions tend to prefer different designs based on national driving conditions and on the regulatory context. The dynamics of patents applied by the main car manufacturers can thus give insights on the competition among technologies for LEVs and on the strategy of firms. In order to apprehend the various technological trajectories examined in the first section (ICEVs, EVs, HVs, FCVs), keywords were identified and used to extract the corresponding patent applications and then to count the number of patents applied by each car manufacturer on each technology. Different keywords were selected: internal combustion engine vehicle (ICEV), diesel engine (DE), fuel cell vehicle (FCV), electric battery vehicle (EBV) and hybrid vehicle (HV). This enables us to examine the cumulated number of patents in each technology per firm as indicative of a firm-specific accumulation of knowledge and advancement of technological trajectory.12 Concerning the dominant design two expressions were used (ICEV and DE) in order to try to isolate the innovations dedicated to diesel vehicles. Indeed as discussed in Section 1, there is a strong debate on the future of advanced diesel vehicles which, according to certain car manufacturers, will become as efficient as hybrid vehicles. So in order to perform the search, three search fields were combined: ‘keywords in title or abstract’, ‘publication date’13 and ‘applicant’. The keywords method offers a simple way to extract a set of potentially relevant patent applications. However, several drawbacks have to be underscored regarding this keywords method. First, the combination of words though linked by the operator AND does not guarantee from finding a patent document that matches the exact query. Indeed typing the keywords in the title or abstract search field enables to find patents having the specified words in the descriptive part even if they are not connected to each other. So the search includes documents in which the invention is described with the various different terms but has nothing to do with what is searched for. This was particularly true for the item ‘electric vehicle’. As a consequence, the number of EV patents proves to be excessively high since it incorporates not only inventions for EV stricto sensu, but also inventions combining electric and electronic devices embarked in a vehicle. Although it is symptomatic from the growing penetration of electric and

10 Frenken et al. [11] show through a patent data analysis that premature lock-in is unlikely to occur. 11 http://ep.espacenet.com. 12 Moreover cumulative distributions of patents can underscore the effects of increasing returns in R&D activity [11]. 13 The publication date is the date on which a patent application was published for the first time, and the applicant is the person or organisation (companies, universities) who filed the patent application.

electronic devices in the car system [17], this creates a bias in the relevant patent set. This is the reason why we chose to narrow down the search for EVs by considering in addition the word ‘battery’ in the title or abstract search field. It resulted in better specifying the sample of relevant patents. Likewise for the item ‘hybrid vehicle’ it is hard to separate the inventions that do concern hybrid cars from those that concern a car solely powered by a combustion engine but including the hybrid term in its description. Related to this drawback, the keywords method makes it difficult to separate different types of innovation. For example, hybrid cars are able to make use of the electricity produced on board for additional functions which gives it a decisive advantage compared to combustion engine cars. Advances relative to the use of electricity produced on board are different from progresses made on the only drive mechanisms of hybrid systems. But the keywords method does not enable to distinguish between these two types of invention. One can also imagine fuel cells applied for supplying individual electronic equipment incorporated in a vehicle and not for powering automobiles. Again the keywords method could not easily separate both innovations.14 A last drawback is linked to the duplication of patents that are subject to application in different countries. So one and the same patent can be counted three times if it has been applied in Japan, in US and in Europe. In spite of this bias, duplicate patents were not eliminated from the sample. As a consequence an innovation which is patented worldwide is counted several times and so is overweighted in our data set. Nevertheless the fact that an innovation is patented worldwide is also a signal on its strategic dimension which can justify an overweighting of this innovation. More generally, the use of patent activities as an indicator of innovation is not without raising problems. First, not all inventions or innovations are patented. So patents give only a partial representation of innovation activities. Second, the patenting activity tends to vary according to the sector and to the point in time under investigation. So depending on the sectors, firms may prefer other appropriation modes and may protect innovations by trade secrets and copyrights instead of patenting. Lastly national differences exist between patent systems in terms of required degree of novelty and constraints attached to the patent system in its whole. For example, it is less expensive to patent in Japan than in the US or in Europe such that it results in an over-representation of Japanese patents.15 However, patent data are a ‘unique resource for the analysis of the process of technical change’, providing an abundant quantity of available information with potential industrial, organisational and technical details [18]. Moreover, patenting is part of the technological responses from some industries, like the automotive one, to deal or to anticipate with environmental regulation. Several studies have shown the correlation, if not the causality, between environmental regulation and patenting. For example Lanjouw and Mody [19] show that increases in environmental compliance costs in United States, Japan and Germany are related to increases in patenting in environmental technology. As well Brunnermeier and Cohen [20] found that innovation, as measured by the number of

14 Nevertheless in order to estimate the relevance of our items, we also tested the intersections between them by counting the number of patents in each possible intersection. This number of patents in intersections reveals to be rather small relatively to our sample of patents. We conclude that our items do separate classes of patents. But we cannot strictly evaluate the relevance of each patent as to its technological content (whether the innovation really concerns the engine technology). For this purpose, relevant patents should then be screened by reading abstracts of patents. 15 Moreover the differences among countries in terms of the required degree of novelty, the flexibility of legislation and the first-to-file or first-to-invent systems also influence the propensity to patent.

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

patents, was consecutive to increases in abatement expenditures which are used as a proxy for policy stringency. Taylor et al. [21] show that patenting activity in SO2 control began after the introduction of a regulatory regime by the 1970 Clean Air Act Amendments and the 1971 New Source Performance Standards. Regarding the automotive industry, Lee et al. [22] show a close relationship between the magnitudes of patenting activity and a series of stringency levels for each of three pollutants considered (HC, CO, NOx), with each increase in stringency leading to increased patenting activity over the period under study (1968–1998). In fact patenting not only reflects the firm’s technological position and its dynamism over time, but is also an umbrella to protect the know-how and the very existence of a firm which could provide bargain power with other partners or guarantee a lead or first-mover advantage if regulation would come to move in the proper direction. This can be related to what Jacob et al. [23] have called lead markets for environmental innovations, that is strict environmental regulations can create lead markets, enabling local firms to benefit from ‘‘early mover’’ advantages and so to export their innovations induced by local market conditions and national regulations. In this perspective, patents can thus represent a strategy to secure inventions potentially able to give an international competitive advantage.

3.2. Results: the diversification of patent portfolios We first present the evolution of the cumulated number of patents applied by our sample of car manufacturers for each technology. Fig. 4 shows a continuous increasing trend for ICEV and DE patents and the progressive increase in HV patents since 1999– 2000, which finally become superior to EVB patents from 2002. This trend illustrates the consequence of the growing involvement of car manufacturers in the development of HV. Regarding FCV, the increase in patents is significant since 2001, but remains relatively low compared to other technologies. The strong involvement of car manufacturers in HV and FCV patents since 2002 and the relative shift in research focus from EV to both HV and FCV is illustrated by the share of car manufacturers in the total number of patents16 for each technology. Indeed, we observe that this share is 57% and 45% for FCV and HV17 in 2005, while it is lower for the other technologies (40% for ICEV, 18% for DE and 36% for EVB in 2005). However, in spite of these advances in LEVs technologies, the figure also shows that the dominant design is still under improvements, in particular, the diesel engine. This is consistent with the fact that the conventional engine is far from being played out since it still represents the core of innovative activities of car manufacturers. The analysis of individual data, such as the share that each car manufacturer represents in the total number of patents applied by our sample of firms (cf. Table 1 in Appendix), calls for three comments. First, Toyota exhibits a position of first-mover regarding both the dominant design (ICEV and DE) and the LEV technologies. This situation leads Toyota to hold a persistent domination over the whole period and to perform the higher share of patents over the set of technologies under consideration, with an exception for FCVs. This leadership illustrates Toyota’s technological accumulation over 15 years both on the dominant design and on the new alternatives. Second, Honda is characterised by an early positioning on ICEV and EVB, but also by a progressive diversification towards DE, FCV and

16 The total number of patents registered for each item, not restricted to our sample of car manufacturers. This number comprises the patents applied by other car manufacturers, by suppliers, by firms from other industrial sectors and by public organisations. 17 For HV this share amounts to 58% in 2003.

207

HV. In 2005, Honda is among the top car manufacturers applying patents on the whole set of technologies. Honda’s and Toyota’s significant shares of patents illustrate the high propensity to patent of Japanese manufacturers compared to the European and the US ones. That is the reason why European and US car manufacturers hold a relatively low share of patents in the sample. It results in an over-representation of Japanese or Korean patents in our sample. In spite of such a leading position of Toyota and Honda, data in Table 1 also clearly show a progressive positioning of all the car manufacturers on FCV and HV especially from the year 2000. In order to examine the competition between the various technologies while handling the problem of Japanese patents overrepresentation, we study separately the patent portfolio of each car manufacturer in the sample. The evolution of such portfolios (cf. Fig. 1 in Appendix) clearly shows their on-going diversification and confirms the previous arguments. Such a diversification illustrates that the evolution of LEVs technologies is characterised by an explorative stage in which firms increasingly widen their patent portfolios. If we look at the patent portfolios in 2005, we observe that all the car manufacturers are involved in the five technologies under study. At first glance, there is no sign of a significant specialisation of car manufacturers even if their portfolios are distributed differently among technologies. In the sense of Stirling [24], we can say that there is an increase in the variety of technologies with differences in the balance of technologies.18 For example, General Motors exhibits one of the most balanced portfolios of the sample in 2005, contrary to Hyundai or Volkswagen. The differences in the balance of portfolios are relevant to capture the strategy of firms. In order to apprehend the dispersion of the portfolio components, we calculate the coefficient of variation of each portfolio through time. Fig. 5 represents the evolution of the coefficient of variation for each firm in the sample. It clearly shows a decrease in the dispersion of portfolios for all car manufacturers. Meanwhile, more or less significant differences persist among firms in the distribution of technologies, as it is particularly the case for GM (low dispersion and thus good balance) and Hyundai (high dispersion and so low balance). The patent portfolios of firms (cf. Fig. 1 in Appendix) confirm the strong dominance of ICEV and DE. Indeed, more than 50% of each patent portfolio is still dedicated to the dominant design (ICEV þ DE): for all the car manufacturers, except for Honda and General Motors, ICEV and DE represent more than 50% of the cumulated number of patents in 2005. What is specific to each car manufacturer is the sharing-out between DE and ICEV: for example in 2005, Toyota, GM and Volkswagen are characterised by a balanced distribution between ICEV and DE while Honda, as well as Renault to a lesser extent, are specialised on ICEV. On the contrary, Hyundai exhibits a portfolio consisting in approximately 63% of DE patents. These results corroborate the continuous improvement of the established technology that goes on exploiting the increasing returns due to its status of dominant design. Moreover, we can also observe, by reading the abstract of patents, that there are more and more environmental innovations in the field of ICEV and DE. Patent data recorded by the French Institute on Intellectual Property (INPI) confirm this statement by showing that 40% of the national automotive patent applications are linked to environmental objectives19 [25]. This feature stresses that the

18 More precisely, according to Stirling [24], the variety corresponds to the number of technological options in the portfolio, while the balance refers to the share of each option in the portfolio. 19 The INPI study [25] also shows that the most part of the patented inventions of the French automotive industry comes to conventional technologies and to the integration of electronics in these technologies. Among these patents, 40% are dedicated to improvements in air admission, 30% to combustion and 30% to exhaust pipe.

208

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

3500 ICEV

3000 2500

DE

2000

FCV

1500

HV

1000

EVB

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

0

1991

500 1990

Cumulated number of patents

Patents applied by the sample of car manufacturers 4000

Publication date

2,5

Index of specialisation in 2005

2

0,5

evb hv fcv icev

HONDAde

TOYOTA

NISSAN

MITSUBISHI

HYUNDAI

evb hv fcv de

NISSAN icev evb

hv de icev

fcv

HONDA 0

2,5

0,5

1

2

1,5

evb hv

VOLKSWAGEN

de

1,5 1

fcv

evb hv

DAIMLER

icev

icev

0,5 evb

PSA 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

fcv

Publication date RENAULT

PSA

DAIMLER

VOLKSWAGEN

hv icev

evb

RENAULT

de

hv

fcv

de

0

evb hv fcv de icev

fcv

de

icev

0,5

1

1,5

2

2,5

Index of specialisation in 2005

1,4 1,2

evb hv

1 GM

fcv

0,8

de

evb hv fcv de icev

icev

0,6 0,4

evb

0,2

hv fcv

FORD

de

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

0

2,5

Index of specialisation in 2005

2

0

evb hv fcv de icev

de

icev

TOYOTA

Coefficient of variation

evb

hv

MITSUBISHI fcv

1

Publication date

Coefficient of variation

de

icev

1,5

0

evb

hv fcv

HYUNDAI

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Coefficient of variation

Fig. 4. Evolution of patents applied by the sample of car manufacturers.

icev

Publication date 0 FORD

0,5

1

1,5

2

2,5

3

3,5

GM

Fig. 5. Evolution of diversity of the patent portfolios (coefficient of variation) and index of specialisation in 2005 for three different groups of car manufacturers.

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

94

19

93

19

92

19

91

19

90

19

19

05

20

20

20

20

20

20

19

19

19

19

19

19

19

19

19

19

04

0 03

0 02

1

01

0,5

00

2

99

1

98

3

97

1,5

96

4

95

2

94

5

93

2,5

92

6

91

3

90

7

95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05

Index of specialisation GM

Index of specialisation TOYOTA 3,5

Publication date

Publication date EVB

95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05

Publication date

Publication date FCV

EVB

19

93

94

19

19

91

02 20 03 20 04 20 05

20

01

00

20

20

99

19

98

19

97

19

96

19

95

19

93

94

19

19

92

91

19

19

90

0

92

0,5

19

1

90

1,5

DE

HV

2 1,8 1,6 1,4 1,2 1 0,8 0,6 0,4 0,2 0

2

ICEV

FCV

Index of specialisation PSA

Index of specialisation RENAULT 2,5

19

DE

ICEV

19

HV

19

FCV

DE

ICEV

209

HV

EVB

DE

ICEV

FCV

HV

EVB

Fig. 6. Evolution of the firm’s specialisation over the period 1990–2005.

established technology tries to compete on the environmental dimension. In terms of old-versus-new competition (cf. Section 2.3), it means that the old technology is also very active in the competition on the environmental function in order to be competitive with the new technologies. All in all, in this context of competition, the established technology tries to compete with the new ones on the environmental function, while the new technologies try to fulfil the genuine function of the dominant design they are supposed to replace. If we focus on LEVs patents, we observe that the share of EVB patents tends to decrease since the end of the 1990s: this is particularly significant for Volkswagen and Honda.20 Toyota is an exception with a constant share dedicated to EVB patents over the period.21 We also notice that the share of HV patents is significantly increasing since 1995 with a first-mover advantage for Toyota in 1992. This result illustrates the argument according to which HV is now considered to be the most promising technology with a significant Japanese leadership. If we compare the portfolios of Japanese firms with the others, we can see that the Japanese ones are characterised by a relatively high share in EVB while the share of FCV is rather low. This is consistent with the fact that the Japanese car manufacturers are positioned on HV

20 In spite of this decrease, Honda is characterised by the highest share of EVB patents in 2005 (25%) in comparison to other manufacturers in the sample. 21 Given the spillovers between EVB and HV, we can suppose that this feature is linked to the leading role of Toyota in HV.

and EVB, since they try to exploit spillovers between technologies and consider that the HV will not only strongly increase its market share in Japan and in US, but also that it is a determining transition step towards FCV. An index of specialisation has been calculated in order to identify the importance of patent applications on each technology for each car manufacturer as regard to the set of firms in the sample. nj

P5 i;t ISPEi;t ¼

j¼1

nji;t

P11 j n ¼ 1 i;t P11 iP 5 i¼1

j¼1

; nji;t

where nji;t represents the number of patents applied by firm i (i ¼ 1,.,11) in technology j (j ¼ 1,.,5) at time t (t ¼ 1990,.,2005). An index superior to one means that firm i is specialised in technology j in comparison with the share of technology j in the whole sample of patents applied by the 11 car manufacturers. The index of specialisation for the year 2005 is represented for the three main regions (US, EU, and Asia). The figures depict different portraits of specialisation adopted by car manufacturers. GM and Renault come out from the sample by the significant number of patents applied on FCV. Hyundai, Mitsubishi and PSA to a lesser extent arise as relatively specialised on DE, whereas Volkswagen, Ford, Daimler-Chrysler and Renault develop a relative specialisation on ICEV. EVB and HV appear to be a specialisation preferred by Asian firms.

210

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

Fig. 6 represents the evolution of such an index for four car manufacturers (Toyota, Renault, GM and PSA) that exhibit specific features as regard to their strategy. As already outlined, Toyota is characterised by a low specialisation (index close to 1), and so a balanced diversification of its patent portfolio, given its early technological accumulation on each technology. Renault appears to be specialised on ICEV (index ¼ 1.55) and FCV (index ¼ 2.06). It is consistent with the recent technological developments followed by the firm in order to exploit the dominant design on the one hand (in particular improvements of the security system and greater fuel consumption efficiency), and to explore new paths on the other hand (in collaboration with Nissan to develop FCV). Concerning GM, in spite of a balanced portfolio, the firm appears to be highly specialised on FCV (index ¼ 2.95). Again it is consistent with the market the firm is targeting, i.e. the US market, and the Californian one in particular, where the legislation in favour of FCV is strongly influential. At last, PSA is still strongly specialised on ICEV (index ¼ 1.07) and DE (index ¼ 1.38) with a continuous increase in specialisation on DE since the end of the 1990s. Such an evolution reflects the technological leadership of PSA on the particle filters and its bet on diesel engines in the future years. Overall, the results demonstrate the diversification of patent portfolios and thus a higher technological diversity at the firm’s level. However, such a diversification is more or less balanced according to the differentiated specialisation of car manufacturers so that a greater diversity among firms can also be outlined. Finally, such differences in the balance of portfolios underscore the differences among countries: while the Japanese leadership on HVs is incontestable, the European car manufacturers are more reluctant to position on hybrid technologies. They still strongly bet on the conventional technology, in particular the diesel one, because they argue that the advanced diesel vehicles will exhibit environmental performances comparable to the ones performed by HVs for a very lower price. Moreover they do not necessarily consider HVs as a compulsory transition step towards FCV.

and FCV, so that an increasing variety of technologies can be observed. However, differences in the balance of technologies as well as differences in the patterns of specialisation adopted by car manufacturers can also be evidenced, thus suggesting different strategic choices among firms. Second, the patent portfolios of firms confirm the strong dominance of ICEVs and diesel vehicles, since more than 50% of the patent portfolio is still dedicated to the dominant design. These results corroborate the continuous improvement of the established technology that goes on exploiting the increasing returns due to its status of dominant design. Meanwhile, the old technology is also very active in the competition on the environmental function in order to sustain the competition with the new technologies. Lastly, the gradual increase in the share of HV patents since 1995 for every car manufacturer, but with a first-mover advantage for Toyota, indicates that HV is now considered to be the most promising technology with a significant Japanese leadership. Hybridising appears to be an efficient medium-term means to generate synergies across the many engine technologies and to try to exploit spillovers and complementarities among technologies. In the future, the progress of LEVs will depend on two factors: the evolution of performance of LEVs (in particular their price and range of use) and the evolution of regulation. According to Maxton and Wormald [6], the evolution of the share of engine technologies in the next 30 years will be characterised by a segmentation of the market. The share of advanced diesel vehicles will still increase while FCV, HV and natural gas vehicles will share less than 30% of the market and remain dedicated to niche markets. But, the evolution of regulation will certainly determine the evolution of each technology. According to the work of Jacob et al. [23] on lead markets, the US could be identified as a lead market for FCV due to the regulatory push in California. The market evolution of FCVs will depend on the diffusion of ZEV standards worldwide, but it seems unlikely that strict ZEV regulation will be introduced as such in other countries. As to HVs, Japan appears as a lead market able to widen to other countries. Finally, advanced ICEVs are supported by the European car manufacturers, especially direct injection technologies for diesel cars have become a market success because of the combination of fuel efficiency and high performances. The broad diffusion of advanced diesel cars will depend on the acceptance of the diesel technology abroad, but also on their ability to meet the strict US environmental standards.

4. Conclusion The diversification of patent portfolios demonstrates that the competition between LEVs is extremely active. First, patent data have shown that all car manufacturers have progressively adopted a position on LEVs, especially on HV

Appendix Table 1. Share of patents applied by each car manufacturer in the total applied by the sample of car manufacturers Publication date

1990

1995

2000

2005

Technology

ICEV

DE

FCV

HV

EVB

ICEV

DE

FCV

HV

EVB

ICEV

DE

FCV

HV

EVB

ICEV

DE

FCV

HV

EVB

Toyota Renault Ford Nissan Honda Mitsubishi Hyundai GM PSA Daimler Volkswagen

0.343 0 0.01 0.105 0.295 0.086 0 0.01 0.029 0.086 0.038

0.266 0.008 0.016 0.282 0.032 0.194 0 0.032 0.016 0.105 0.048

1 0 0 0 0 0 0 0 0 0 0

0 0 0 1 0 0 0 0 0 0 0

0.593 0 0.037 0.111 0.111 0.111 0 0.037 0 0 0

0.271 0.035 0.101 0.114 0.173 0.172 0.002 0.015 0.034 0.056 0.027

0.328 0.011 0.016 0.248 0.012 0.264 0.005 0.019 0.016 0.062 0.02

0.357 0 0 0.071 0.071 0.071 0 0 0 0.429 0

0.283 0.011 0.109 0.098 0.043 0.228 0 0.011 0.011 0.141 0.065

0.284 0.011 0.03 0.181 0.246 0.149 0.003 0.011 0.014 0.049 0.024

0.265 0.05 0.083 0.12 0.199 0.127 0.003 0.013 0.019 0.087 0.035

0.321 0.011 0.02 0.243 0.009 0.232 0.041 0.015 0.016 0.071 0.022

0.269 0.03 0.03 0.06 0.075 0.09 0 0.045 0 0.373 0.03

0.358 0.043 0.026 0.238 0.154 0.098 0.008 0.016 0.003 0.045 0.013

0.242 0.023 0.014 0.24 0.236 0.133 0.035 0.01 0.009 0.044 0.015

0.269 0.068 0.098 0.101 0.172 0.093 0.004 0.019 0.027 0.098 0.053

0.268 0.028 0.031 0.221 0.012 0.191 0.105 0.02 0.035 0.06 0.03

0.189 0.091 0.035 0.262 0.166 0.024 0.029 0.067 0.017 0.084 0.036

0.287 0.037 0.08 0.214 0.195 0.068 0.033 0.024 0.021 0.031 0.01

0.242 0.023 0.038 0.225 0.215 0.123 0.054 0.015 0.012 0.039 0.013

20%

EVB

20%

Publication date

Publication date EVB

ICEV

20%

Publication date

Publication date EVB

ICEV

20%

Publication date

Publication date EVB

HV

FCV

DE

ICEV

EVB

HV

FCV

DE

ICEV

2004

0% 2002

2004

2002

2000

1998

1996

1994

1992

0%

40%

2000

20%

ICEV

60%

1998

40%

DE

80%

1996

60%

FCV

Nissan patent portfolio

1992

80%

HV

100%

1990

Hyundai patent portfolio

1994

DE

Share of cumulated patents

FCV

HV

100%

1990

Share of cumulated patents

EVB

2004

0% 2002

2004

2002

2000

1998

1996

1994

1992

0%

40%

2000

20%

ICEV

60%

1998

40%

DE

80%

1996

60%

FCV

Ford patent portfolio

1992

80%

HV

100%

1990

Honda patent portfolio

1994

DE

Share of cumulated patents

FCV

HV

100%

1990

Share of cumulated patents

EVB

2004

0% 2002

2004

2002

2000

1998

1996

1994

1992

0%

40%

2000

20%

ICEV

60%

1998

40%

DE

80%

1996

60%

FCV

Volkswagen patent portfolio

1992

80%

HV

100%

1990

Renault patent portfolio

Share of cumulated patents

ICEV

1994

DE

100%

1990

Share of cumulated patents

FCV

HV

2004

Publication date

Publication date EVB

2002

0% 2000

2004

2002

2000

1998

1996

1994

1992

0%

40%

1998

20%

60%

1996

40%

80%

1994

60%

General Motors patent portfolio

1992

80%

211

100%

1990

Share of cumulated patents

Toyota patent portfolio 100%

1990

Share of cumulated patents

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213

Mitsubishi patent portfolio 100% 80% 60% 40% 20%

2004

2002

2000

1998

1996

1994

1992

0%

1990

Share of cumulated patents

212

Publication date HV

FCV

DE

ICEV

Peugeot patent portfolio 100% 80% 60% 40% 20%

2004

2002

2000

1998

1996

1994

1992

0%

1990

Share of cumulated patents

EVB

Publication date HV

FCV

DE

ICEV

Daimler-Chrysler patent portfolio 100% 80% 60% 40% 20%

2004

2002

2000

1998

1996

1994

1992

0%

1990

Share of cumulated patents

EVB

Publication date EVB

HV

FCV

DE

ICEV

Fig. 1 Patent portfolio of car manufacturers.

References [1] Arthur WB. Competing technologies: an overview. In: Dosi G, Freeman C, Nelson R, Silverberg G, Soete L, editors. Technical change and economic theory. London: Pinter; 1988. p. 590–607. [2] Arthur WB. Competing technologies, increasing returns, and lock-in by historical events. Economic Journal 1989;99:116–31. [3] Kemp R. Zero emission vehicle mandate in California: misguided policy or example of enlightened leadership?. In: Sartorius C, Zundel S, editors. Time strategies, innovation and environmental policy. Cheltenham, UK, Northampton, MA, USA: Edward Elgar Publishing; 2005. [4] Pilkington A. The fit and misfit of technological capability: responses to vehicle emission regulation in the US. Technology Analysis and Strategic Management 1998;10(2):211–24.

[5] European Commission. Chapter 4: the European automotive industry. In: European competitiveness report 2004. Commission Staff Working Document, SEC, European Communities; 2004, 1397. [6] Maxton GP, Wormald J. Time for a model change. Cambridge University Press; 2004. [7] ERTRAC. European road transport research advisory council. Strategic Research Agenda December 2004. [8] Hoed RVD, Vergragt PJ. Technological shifts and industry reaction: shifts in fuel preference for the fuel cell vehicle in the automotive industry. Paper presented at ‘Knowledge and economic and social change: new challenges to innovation studies’ Conference, Manchester, UK; April 2003. [9] Hoed VDR. Sources of radical technological innovation: the emergence of fuel cell technology in the automotive industry. Journal of Cleaner Production 2007;15:1014–21.

V. Oltra, M. Saint Jean / Journal of Cleaner Production 17 (2009) 201–213 [10] Well-to-Wheels Report. Well-to-wheels analysis of future automobile fuels and powertrains in the European context, Concawe, Eucar, JRC-EC, version 1b; January 2004. [11] Frenken K, Hekkert M, Godfroij P. R&D portfolios in environmentally friendly automotive propulsion: variety, competition and policy implications. Technological Forecasting and Social Change 2004;71:485– 507. [12] Hekkert M, Van den Hoed R. Competing technologies and the struggle towards a new dominant design: the emergence of the hybrid vehicle at the expense of the fuel cell vehicle. Greener Management International 2006; (47)::29–47. [13] Christidis P, Hernandez H, Goergakaki A, Peteves SD. Hybrids for road transport – status and prospects of hybrid technology and the regeneration of energy in road vehicles. IPTS-JRC report; February 2005. [14] Oltra V, Saint Jean M. The dynamics of environmental innovations: three stylised trajectories of clean technology. Economics of Innovation and New Technology 2005;14(3):189–212. [15] David PA. The economics of QWERTY. American Economic Review 1985;75: 332–7. [16] Sartorius C, Zundel S, editors. Time strategies, innovation and environmental policy. Cheltenham, UK, Northampton, MA, USA: Edward Elgar Publishing; 2005.

213

[17] Lung Y. Coordinating competencies and knowledge in the auto industry. Special issue of International Journal of Automotive Technology and Management 2004;4(2/3). [18] Griliches Z. Patent statistics as economic indicators: a survey. Journal of Economic Literature 1990;28:1661–707. [19] Lanjouw J, Mody A. Innovation and the international diffusion of environmentally responsive technology. Research Policy 1996;25:549–71. [20] Brunnermeier SB, Cohen MA. Determinants of environmental innovation in US manufacturing industries. Journal of Environmental Economics and Management 2003;45:278–93. [21] Taylor MR, Rubin ES, Hounshell DA. Effect of government actions on technological innovation for SO2 control. Environmental Science and Technology 2003;37:4527–34. [22] Lee J, Veloso F, Hounshell DA, Rubin EA. Technology-forcing regulation and innovation: a panel data study in automotive emissions control. Under review at Research Policy; 2005. [23] Jacob K, Beise M, Blazejczak J, Edler D, Haum R, Ja¨nicke M, et al. Lead markets for environmental innovations. Germany: Physica-Verlag Heidelberg; 2005. [24] Stirling A. On the economics and analysis of diversity. SPRU Working Paper No. 28; 1998. [25] Inpi-Opi. Automobile et environnement, Etude du de´partement des brevets de l’INPI; 2006.