You are on page 1of 8

World Patent Information 24 (2002) 512 www.elsevier.

com/locate/worpatin

The electric vehicle: Patent data as indicators of technological development


Alan Pilkington *, Romano Dyerson, Omid Tissier
School of Management, Royal Holloway, University of London, Egham, Surrey, TW20 0EX, UK

Abstract This paper reports a study into the use of US patents to analyse responses to regulatory change in the automobile industry. Conrming that patents are a rich indicator of technological development, it focuses on the development of the electric vehicle (EV) and, in particular, the identication of networks of rms developing EVs. A key nding of the study is the way that car rms have formed links with competitors and with rms and inventors outside the automobile industry, in order to develop this technology. This contradicts the normal product development methods in the automobile industry, and is attributed to the substantially dierent technologies required for EVs. In addition, the limitations of dening patent searches in terms of productsrather than technologiesare discussed. 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Electric vehicle; Environmental regulations; Technological development; Patent classication; Patent content analysis; Patent citation analysis

1. Introduction The use of patent information is gaining increasing attention in the elds of innovation and technology management. Patent data represent a valuable source of information that can be used to plot the evolution of technologies over time. More recently, Ernst [10,18] has developed a methodology to identify and investigate the role played by individual rms, and indeed key individuals, acting as primary sources or more particularly as gatekeepers of emerging technologies. The changing regulatory environment, principally in California but more generally elsewhere, suggests a shift in the hitherto comparatively stable technology environment of automotive manufacturers. This is a shift from a proven sustaining technology to an unproven disruptive technology [2,12] namely the electric vehicle (EV). With little or no experience of the systems needed to develop a viable EVnotably batteries, electric motors and electronic controllers (see Table 1)the traditional car makers have had to enlist the support of many external rms outside the boundaries of the traditional automotive manufacturing and supply environments.

These new relationships have forced an evolution of the new product development process in the automotive sector as a result of developments in the supply chain [8]. 2. Methodology We identied 268 US patents in the international classication code (IPC) B60L 11/- using the esp@cenet system of the European Patent Oce (EPO). The database lists most current US patents going back to 1976 and was used to identify the patents for the study as it gave more results than using the US Patent and Trademark Oces (USPTO) own search interface. However, there are some reservations about using the esp@cenet system, as there are no precise statements about what exactly is covered in terms of timescales and country, and as a result the EPO makes no guarantees about its completeness or accuracy. The B60L11/- IPC code used represents Electric propulsion with power supplied within the vehicle, and was used in preference to the US patent classication code (UPC) 180/65.1 which covers Vehicles wherein an electric motor in the body or on the body-frame drives the vehicle as it gave more results. One of the aims of this study was to investigate whether this IPC code based search strategy was sucient to provide useful results and so we were interested in capturing as many

Corresponding author. Tel.: +44-1784-443790. E-mail address: a.pilkington@rhul.ac.uk (A. Pilkington).

0172-2190/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 1 7 2 - 2 1 9 0 ( 0 1 ) 0 0 0 6 5 - 5

A. Pilkington et al. / World Patent Information 24 (2002) 512

Table 1 EV system changes and technology adoption Vehicle sub-system Body group Body-in-white Paint/coatings Glass Body trim Seats Instrument panel HVAC Engine group Base engine Engine control Transmission group Transaxel Transmission controls Chassis group Frame Suspension Steering Brakes Fuel storage Refueling Final drive Wheels/tyres Bumpers Fluids Changes to existing design for battery electric vehicle (BEV) introduction Lightweight construction Amend to body material Replace with lightweight Redesign for lightweight Redesign for lightweight Changed functions Redesign for electrical operation Replace with electric motor System controller Extensively modied Deleted Replace with lightweight Redesigned for change in weight All wheel drive adopted Redesigned for new weight Tank replaced with battery compartments in frame Replaced with onboard charger Replaced with electric motor Low rolling resistance Amend to suit body and weight Many removed with drive-train New technology used

Space-frame and composites In mould coatings Polycarbonate construction New plastics and methods Plastic frames Complex system monitoring Self contained, not powered by engine Advanced AC motors Power controller and regenerative braking systems Composite/ceramic materials

Space-frame and composites Plastics and composites Plastics and composites Regenerative electronics Depends on battery technology adopted Electronic inductive charging system Advanced AC motor New materials and proles Plastics and composites

Source: Dyerson and Pilkington [8].

patents for analysis as possible. As we explore in the discussion, there are problems with following patent classication systems when identifying data sets to analyse particular product trajectories. One potential problem to recognise here is that the USPTO naturally uses its own classication system for patent searches and so applying IPC terms given by the USPTO invariably suers from a lack of the detailed appreciation that results from the continuous feedback examiners receive from classifying and searching documents using their own system. This is compounded by the diering philosophies underlying the UPC and IPC systems. Full contents of these patents were then collected using the online data sets of the US Patent Oce. American data is preferred for the comprehensiveness of their data content [1,3] and their applicability as the centre of attention on EV development in response to air quality regulations. The patent information was coded and analysed within the spreadsheet package Excel to investigate the various elds of information available. This coded patent data included: the date of issue allowing various chronologies to be developedauthor and assignee informationallowing the identication of gatekeepers, rm portfolios and activity rankings. Details of cited patents were used to identify key inventions, rank patent quality and allow the analysis of activity clusters.

It is important here to remember that many of the patents analysed originated in dierent countries and so are developed using slightly dierent approaches. One example is the way that citation practices dier between the US and most other territories. This is primarily due to the law in the USA where the validity of a granted patent can, and often is, challenged on the basis that the applicant has not made a full and fair disclosure of the prior art. As a result, the cited patents of a US patent often exhibit a bias toward those of the applicants themselves, and are also notoriously noisy, simply because the attorney is trying to ensure that all angles are covered. The nature of the patents identied above was analysed using a rudimentary content analysis that involved classifying each patent depending on the nature of any application and the technology reported within the abstract. This was performed to see if the patents captured represented a valid and complete picture of technology development for the EV.

3. Rise of the electric vehicle Signicant developments in emission regulation are presently forcing rms towards new product programmes that utilise technologies positioned outside of

A. Pilkington et al. / World Patent Information 24 (2002) 512

their normal frame of reference. The evolving regulatory regime, particularly in California but more generally elsewhere, suggests a shift in the hitherto comparatively stable technology environment of automotive manufacturers. That is, a shift from a proven sustaining technology to an unproven disruptive technology [2,12]. More specically, changing regulatory requirements are stimulating a derived but uncertain demand for cleaner cars. From the point of view of the traditional car makers, demand uncertainty is exacerbated by technological uncertainty because existing technological competencies and organisational knowledge sets based on the mature internal combustion engine (ICE) are unlikely to meet the regulatory requirements. And yet failure to come to terms with EV technology, particularly given the regulatory push, may open the established car makers to threats from new entrants [7]. Table 1 provides some details as to range of new technologies used in the construction of EVs prototypes, over and above existing technologies in use for ICEs. The most evident regulation to prompt or encourage this change in technology has been the California Air Resources Boards (CARB) zero-emission vehicle (ZEV) mandate. This regulation, primarily introduced to cut emissions, stipulated in its earliest form, before being fundamentally revised, that manufacturers selling more than 35,000 vehicles in California should have 2% of their sales made up of ZEV during 1998. This in eect mandated the introduction of a large number of BEVs to the US market, as this was then the most suitable technology, prompting further development work. A particular technological problem of interest here is the nature of the development programme in producing BEV prototypes, which in eect represent a complex product system (CoPs). That is high cost, engineeringintensive products, systems, networks and constructs [13]. These are highly customised, one-o or small batch products that in their manufacture typically involve the complex interaction of a range of production technologies (see for example Refs. [6,13,14]). The problem for conventional car manufacturers is that their previous accumulated experience in designing and mass-producing automobiles based on the ICE may be ill-suited to the technologies required in producing EVs. Indeed, as suggested by Table 1, a new set of skills and building blocks are required for successful development and commercialisation. Such pressures appear to be pushing incumbents into seeking strategic alliances and acquisitions in an eort to bolster their existing set of technological know how. One potential problem with this approach is, as Cohen and Levinthal [5] have identied, absorptive capacity. That is, the rms ability to identify, assimilate, and exploit know-how type knowledge from its environment depends upon its willingness to engage in a substantial amount of in-house R&D. A further problem is suggested by

Ernsts [9] nding that the technological performance of inventors tends to be highly concentrated, making particular inventors highly inuential in technology development. At a time of changing technological know-how, this potentially exposes existing car makers to dependence on outside experts and gatekeepers. Put another way, incumbents have been investing resources in the prevailing dominant design. Over the past century, the automobile industry has evolved from a multiplicity of competing designs to a gradual convergence on a common approach or dominant designthe ICE. That convergence has been marked by the transition from an early uid stage, dominated by small innovative rms, to an industry overseen by a handful of large established rms. Maturity is reected in the shift of competitive emphasis from product characteristics to production process [11]. Using Rothwell and Gardiners [16] terminology, the dominant design has proven to be very robust (i.e. market established) producing a series of design families from Fords model T to the Focus of today. Regulatory change however is threatening to make the existing technology portfolios of the dominant car makers increasingly redundant. This has prompted signicant new investment by the dominant carmakers, working in tandem with outside partners. In the US the big three car producers (General Motors, Ford and Chrysler) have all been involved in pilot schemes and development programmes, with General Motors perhaps having the upper hand as a result of its EV1 programme [17]. Ford and Chrysler have similarly experimented with EV through their Ranger and EPIC models respectively. All three rms can be seen as experimenting with outside partners in an attempt to expand their own knowledge bases. While not a focus in this paper, most European and Japanese automobile manufacturers face the same restrictions from the CARB ZEV mandate, and so have also been developing EVs for both indigenous and export use (see for example, Refs. [15,19]). Toyota and Honda have developed BEVs which have entered production in small numbers, and their experiences in electric drives have clearly played a part in the development of the widely available Prius and Insight hybrids. Regulatory change has also provided the opportunity for new entrants into the sector. Away from the existing carmakers some other rms have started exploring EV manufacturing. For example the specialist Swiss plastics maker Horlacher has developed a range of BEVs utilising lightweight carbon bre bodies. Similarly, PIVCO of Norway, which was recently taken over by Ford to create TH!NK, produced several light weight EVs, including the CITI/City Bee which was being used in the Bay Area Rapid Transport (BART) and Pacic Gas and Electric Companys station car project. The project, and others like it, involves providing recharging bases at

A. Pilkington et al. / World Patent Information 24 (2002) 512 Table 2 Signicant EV patent owning companies Assignee company Toyota Jidosha Kabushiki Kaisha General Electric Company Mitsubishi Jidosha Kogyo KK Ford Motor Company Nippondenso Co., Ltd. Nissan Motor Co., Ltd. Aisin Aw Co., Ltd. Daimler-Benz AG Chrysler Fuji Electric Co., Ltd. General Motors Corporation Honda Giken Kogyo Kabushiki Kaisha Lucas Industries Limited Newport News Shipbuilding and Dry Dock Company USA Government ABB Brunswick Corporation Isuzu Ceramics Research Institute Co., Ltd. Kabushikikaisha Equos Research Kvaerner ASA Mannesmann Aktiengesellschaft Nippon Soken, Inc. Seiko Epson Corporation Shinko Electric Co., Ltd. SMH Management Services AG Tokyo Shibaura Electric Company, Ltd. Westinghouse Electric Corp. Zebco Corporation Number of patents held 24 14 5 4 4 4 3 3 3 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2

transit stations and 55 mph65 mile range BEVs for commuters to make the journey from home to the station. A number of small rms are also engaged in designing and producing EVs either from the drawing board or by converting existing ICE vehicles. These rms specialise in selling converted vehicles and conversion kits to the general public and eet operators. One example, Solectria of Waltham Massachusetts, is involved in a deal exporting vehicles to Japan through the Sanoh Industrial Company Ltd. These vehicles, such as the Force sedan and E-10 pick-up, were developed with help from various national and local government backed projects. The products and programmes outlined above have involved the generation of networks of rms, bringing in expertise from outside the traditional automotive sector. For example, the US big three have entered agreements with specialist battery manufacturers such as Ovonics and established new divisions to develop other vehicle systems such as charging equipment. Similarly, Toyotas RAV4-EV led to close links with electrical giant Matsushita for batteries and GMs Delco for charging technology. Honda, who have long been at the forefront of developing advanced ICE engines, have been working with the US utility Pacic Gas and Electric Company in preparation for the ZEV regulations, whilst Nissan have a design unit in San Diego and make use of batteries developed by Sony.

4. Results and discussion We found that the leading proponents of EV technology were Toyota and General Electric, not surprising given the new hybrid product from the Japanese rm and the involvement of General Electric in General Motors EV1 programme. Table 2 details the rms involved in the EV patents, ranked by the number of patents each holds. Our results support ndings from other studies [8] that a network of rms, many outside the existing automobile new product development process, is developing the EV, and as such supports the use of patent data to examine technology development. By way of comparison, we repeated the same analysis but adopting a patent class dominated by traditional ICE technology, B60K5/04 Arrangement or mounting of ICE with the engine main axis, e.g. crankshaft axis, transversely to the longitudinal centre line of the vehicle. The results, presented in Table 3, support the conclusions of other studies showing the focus of traditional automotive new product development activities to be within the existing supply base [4]. This highlights the signicance of BEV new product development activities as a shift away from the standard activities within the auto industry.

Table 3 Companies patenting traverse engine designs (B60K5/04) Assignee company Mazda General Motors Yamaha Honda Toyota Unassigned Citroen Daimler Chrysler Nissan Ford Porsche Torotrak (GB) Stockamolen AG Number of patents held 15 4 4 4 3 3 2 1 1 1 1 1 1

The chronology of EV patenting activity was particularly striking with various cycles of activity evident. These did not align with measures of patenting activity reported by the US PTO (see Fig. 1). We explored whether these dierences observed in the gure were signicant by performing a v2 (chi-squared) test. This statistic is the standard method for determining if a given sample (the EV data) exhibits the same distribution as a known population (all US patents granted). The calculated v2 comparison statistic was 48.784, which exceeds the critical value of 38.932 at the 0.01 condence

A. Pilkington et al. / World Patent Information 24 (2002) 512

Fig. 1. Chronology of the number of EV patents granted (compared to all US patents granted).

level, suggesting that the distribution of B60L 11/- patents is signicantly (99.99% likelihood) inuenced by factors other than general patenting behavior. It is conjecture at the moment and further work is needed to examine this phenomenon, but an impression is that these cycles might show a relationship between technology development activity and other factors such as the passage and development of emission regulations, oil price uctuations and government research initiatives.

Turning now to consider the rms engaged in EV development, Fig. 2 provides a chronology of key patents cited in category B60L 11/- (on the left) and also identies the rms citing those key patents (on the right). The citing rms are separated for clarity into a number of categories: the top three citing car makers (Toyota, Ford and Nissan); other car makers (this includes Fiat, General Motors, Isuzu, Mitsubishi and Peugeot); car component companies (eight companies

Fig. 2. Relationships between citing rms and key patents.

10

A. Pilkington et al. / World Patent Information 24 (2002) 512 Table 4 Type of vehicle application identied in the patents Marine vehicle 22% Road vehicle 36% Other vehicle type 10% No vehicle type specied 32% Total 100%

in total); and others (eleven in total). The lines in the gure indicate the key patents that the companies cited in their own applications. For the sake of clarity, independent inventors citing key patents have been excluded from Fig. 2. Here we dene key to be any granted patent that attracted ve or more subsequent citations [9]. We should note with caution the lower frequency of citations as the modernity of the patent increases; this is to be expected given that there is likely to be a time lag for new patent information to diuse within the community. However, it is striking that so few of these 24 key patents are held by American carmakers. Indeed, neither Ford nor Chrysler appears as key patent holders in the Figure, although General Motors 1973 patent appears to be quite heavily cited. Instead many of the key patents appear to be held by Asian carmakers, together with a small European presence. This may be reective of wider trends. Levin et al. for example found patenting to be comparatively unimportant as an appropriation device in the automobile sector [20], although this nding related to the mature and presumably diused technology found in ICEs rather than the youthful EVs. Going further, judged by number of citations, most of the heavily cited patents are held either by non-American rms or by independent inventors, with the exception of General Motors 1973 patent. A slightly richer pattern emerges when examining the rms engaged in citing the key patents, as can be seen in Fig. 2. The gure shows the leading individual citers are a mixture of Japanese and American carmakers, notably Toyota, Ford and Nissan. Toyota in particular appears to be very active in the area with the rm accounting for more than 25% of all citations made. Beyond this group, other car companies are clearly underrepresented, accounting for only a few citations. Japanese and European component makers dominate the second most signicant group of patent citers with non-car rms and research institutes close behind (see also Table 2). This nal group of rms are of particular interest, as they represent rms from outside the traditional new vehicle product development networks and is representative of the range of new technologies that BEV development utilises. We do need to be aware again that there are dierent patent practices that arise in dierent territories. It is accepted that Japanese and European rms are often more frequent citers of key patents because the citations are primarily provided by a third party (e.g. as a result of a European patent search), and therefore more likely to represent the closest art, rather than the possibly more comprehensive, but much less selective approach in the US. Another signicant nding from this study has been the importance when using patent data, of phrasing the research questions in terms of technologies and innovations rather than products and uses. Investigating the

development of the EV was achieved using the patent grouping B60L 11/-, and the ndings in Table 2 suggest that we have managed to capture a picture of the rms involved in the development of the technology, but the patents included within this classication relate to many other applications apart from EVs. A brief content analysis was performed on the patents by classifying each one against which type of vehicle the patent abstract discussed. The results, shown in Table 4, show how poor the IPC term chosen was in identifying the automobile BEV typeour original intentionbecause the denition of this specic group embraces a wide range of EV not just automobiles. We are also aware that the data set excluded patents representing the key sub-systems needed for EV introduction identied in Table 1. A potentially high number of patents related to hybrid technologies when originally we expected battery power to dominate as it has the demonstration vehicles and those available in the market to date. This aspect was investigated using another content analysis, this time focusing on the sub-system utility aspect of the patent rather than the vehicle type. The results, shown in Table 5, reect the ndings elsewhere that we have managed to obtain a picture of the technology development, but not the whole story. There were no patents dealing with body design or materials and few relating to alternative battery technologies themselves, as most with a battery focus concerned the incorporation of any type of battery technology within a vehicle. These concerns about the nature of patenting are an area of further work, and one solution is to extend the survey by using patents returned by searching for the phrase electric vehicle. Whilst this method will return a wider range of sub-technologies used in EV development and require more manual selection, it will still tend to exclude disruptive technologies that have yet to be earmarked for EV use. Furthermore, expanding the coverage of sub elements important to the EV (albeit with limitations), upping in eect the number of identied patents from 268 to 3,306, increases the susceptibility of the raw data set to noise problems as innovations of little relevance to the EV are inevitably included. This highlights the problem of trying to use patents to analyse the development of a product rather than a technology. For a patent to meet the requirements of the relevant patent legislation, it necessarily focuses on the technical innovation itself; the applications of that technology may be specied in some detail, but may some-

A. Pilkington et al. / World Patent Information 24 (2002) 512 Table 5 Sub-system focus of the patents Control system 20% Monitoring system 1% Auxiliary system 20% Power system 24% Charging system 9% Hybrids 11% Batteries 15% Total 100%

11

times be much less explicit. The normal way for patent searchers to perform such an expansion is by building lists of related technologies and incorporating the data from other patent classes. For example, our own set can be enlarged by adding the classes relevant to Table 1, e.g. HVAC is covered by B60H1, structures by B62D21 B62D29, hybrid vehicle control is in B60K, electric motors and controllers are in H02K/H02P, etc. The advances captured in this way may, or may not, be used in EV development and this is a major limitation to utilising patent data to explore product advances.

Acknowledgements Our thanks to the referees for their comments.

References
[1] Albert M, Avery D, Narin F, McAllister P. Direct validation of citation counts as indicators of industrially important patterns. Res Policy 1990;20(3):2519. [2] Bowyer JL, Christensen CM. Disruptive technologies: catching the wave. Harvard Business Rev 1995;73(1):4353. [3] Brockho K. Instruments for patent data. Technovation 1992; 12(1):4518. [4] Clark K, Fujimoto T. Product development performance: strategy, organization, and management in the World Auto Industry. Boston: Harvard Business School Press; 1991. [5] Cohen W, Levinthal DA. Innovation and learning: the two faces of R&D. Econ J 1989;99:56996. [6] Davis A. The life cycle of a complex product system. Int J Innovation Mgmt 1997;1(3):22956. [7] Day GS, Schoemaker PJH. Avoiding the pitfalls of emerging technologies. California Mgmt Rev 2000;42(2):833. [8] Dyerson R, Pilkington A. Innovation in complex systems: regulation and technology towards the electric vehicle. Int J Innovation Mgmt 2000;4(1):3349. [9] Ernst H. Key inventors: implications for human resource management in R&D. In: Kocaoglu D, Anderson T, editors. Technology and innovation management. Portland: PICMET, 1999. p. 4207. [10] Ernst H, Vitt J. The inuence of corporate acquisitions on the behaviour of key inventors. R&D Management 2000;30(2):105 19. [11] Froud J, Haslam C, Johal S, Williams K. Breaking the chains? A sector matrix for motoring. Royal Holloway School of Management, unpublished. [12] Henderson RM, Clark KB. Architectural innovation: the reconguration of existing product technologies and the failure of established rms. Administrative Sci Quart 1990;35:930. [13] Hobday M. Product complexity, innovation and industrial organisation. Res Policy 1998;26:689710. [14] Miller R, Hobday M, Leroux-Demers T, Olleros X. Innovation in complex systems industries: the case of ight simulators. Industry and Corporate Change 1995;4(2):363400. [15] Quandt CO. Manufacturing the electric vehicle: a window of technological opportunity for Southern California. Env Plann A 1995;27:83562. [16] Rothwell R, Gardiner P. Design and competition in engineering. Long Range Plann 1984;17(3):7891. [17] Shnayerson M. The car that could: the inside story of GMs revolutionary electric vehicle. New York: Random House; 1996. [18] Teichert T, Ernst H. Assessment of R&D collaboration by patent data. In: Kocaoglu D, Anderson T, editors. Technology and innovation management. Portland: PICMET, 1999. p. 7886.

5. Conclusion The use of patent information is gaining increasing attention in the elds of innovation and technology management. Patent data represent a valuable source of information that can be used to plot technological evolution over time. However, there is evidence from this study that particular care is required when establishing the data set for a patent analysis. Patents, because of their nature, represent technological advances and include speculation about how that technology may be used in the future. As a result the researcher must respect that they do not necessarily make clear statements about how the patent will be implemented for specic products. The patent author is often trying to cover as many future possible uses for the innovation as they can. In this particular case, a limited survey of patent data using the international classication scheme has allowed an insight into the development of new networks as a result of the technological development required for EVs to become viable product propositions. The work has highlighted that the technological advances are not just in developing the vehicle itself, but also in the sub-systems that make up the whole package. Whilst this is not unknown in the traditional approach to new car development [4], EV development has many fewer existing power-train or mechanical components to design the vehicle around and so the new product development process has to innovate these systems as well as repackage modied existing systems. The real diculty for the car makers is their limited knowledge of electric power-trains and so we see a reliance on rms beyond the boundaries of their existing new product development networks.

12

A. Pilkington et al. / World Patent Information 24 (2002) 512 global industry, and examining the adoption of environmental management systems. Dr. Romano Dyerson is an Economist, and previously a Research Fellow at London Business School. His long-standing research interests in technology and innovation include the evolution of strategic technologies at the rm and sector level. Omid Tissier is studying for a BSc in management studies with economics. His main interests lie in Financial Accounting.

[19] Toyota S. The electric vehicle: the challenge of the next century. Report of EVS 13, Osaka, Electrifying Times 1996;4(1). [20] Levin R, Klevorick W, Nelson R, Winter S. Appropriating returns from industrial research and development. Brookings Papers on Economic Activity 3, Washington, 1987. Dr. Alan Pilkington is an Engineer, and formerly a manufacturing Manager for a UK automobile producer. He is a Senior Lecturer at Royal Holloway, University of London. He is currently working on the impact of changes in technology and regulation on the structure of

You might also like