You are on page 1of 13

ARTICLE IN PRESS

Int. J. Production Economics 118 (2009) 473485

Contents lists available at ScienceDirect

Int. J. Production Economics


journal homepage: www.elsevier.com/locate/ijpe

Managing technology for highly complex critical modular systems: The case of automotive by-wire systems
Christian Coronado Mondragon a,, Adrian Coronado Mondragon b, Roger Miller a, Etienne Coronado Mondragon c
a Department de Mathematiques et Genie Industriel, Ecole Polytechnique de Montreal, Campus de lUniversite de Montreal, 2900, boul. Edouard-Montpetit, Montreal, QC, Canada b Centre for Logistics Research, The Logistics Institute, The University of Hull Business School, Cottingham Road, Hull HU6 7RX, UK c Department of Electrical and Computer Engineering, Universite de Sherbrooke, 2500 Boul. De LUniversite, Sherbrooke, QC, Canada

a r t i c l e in fo
Article history: Received 10 December 2007 Accepted 20 December 2008 Available online 20 January 2009 Keywords: Modular systems High complex industries Automotive suppliers By-wire systems Managing complexity

abstract
The motor vehicle industry has experienced the adoption of several new complex technologies that affect the approach with which OEMs, suppliers tackle product complexity. Nonetheless, the implications of these complex technologies have the potential to change the way automotive OEMs and suppliers relate to each other. This paper addresses the fact that technological specialization produces a complex electronic/electric vehicle architecture; which creates an interdependency among all members of the automotive innovation networks, a fact that has not been addressed in existing literature. For this in-depth case study, the specic examples of brake-by-wire and steer-by-wire systems were selected as subjects. & 2009 Elsevier B.V. All rights reserved.

1. Introduction The management of product complexity traditionally has being tackled by academics through the extensive use of modular product architectures (Camuffo, 2000; Garud et al., 2003; Langlois and Robertson, 1992; Fixson and Veloso, 1997). Furthermore, Carliss et al. (1997) contend that modularity facilitates the creation of complex product architectures by developing subsystems that can be designed independently and still work as a whole. Automotive systems such as automobiles, heavy-duty trucks, coaches and transit buses are complex architectural systems which required the interaction of many components to work in synchrony. Thus, modularity permits the management of complexity of products to
Corresponding author.

E-mail addresses: Christian-etienne.coronado-mondragon@polymtl.ca (C.C. Mondragon), a.coronado-mondragon@hull.ac.uk (A.E. Mondragon), roger.miller@polymtl.ca (R. Miller), etienne.coronado@usherbrooke.ca (E.C. Mondragon). 0925-5273/$ - see front matter & 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.ijpe.2008.12.015

attain sustainable growth (Christensen, 1997, in Dorf; Baldwin and Clark, 1997, 2000; Hsuan, 1999, 2003; Sanchez and Mahoney, 2003). According to Baldwin and Clark (1997) the successful implementation of a modular architecture depends upon key factors; such as, the architecture denition, the interface between coremodule denition and standard modules. Thus modularity creates ease of integration and is also an effective tool to increase exibility (Baldwin and Clark, 1997, 2000; Garud and Kumaraswamy, 1995; Garud et al., 2003). The works from Doran et al. (2007), Fixson and Sako (2001), Fixson and Veloso (2001), Helper et al. (1999), Holweg and Pil (2004), Miller (1994), Sako (2003) Takeishi and Fujimoto (2003), among many others have identied automotive systems as modular systems. Helper et al. (1999) and Sako (2001) foresee, for automotive product architecture, the adoption of heavily modularized vehicles that are linked together by the effectiveness of the architectural interfaces. Specically, Helper et al. (1999) propose that critical modular vehicle subsystems are produced by the OEMs, outsourcing

ARTICLE IN PRESS
474 C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485

non-modular components. Modularity, as a strategy for value creation, has been acknowledged as a systemic innovation (Nystrom, 1990; Birchwall and Green, 2006), consequently autonomous innovations are held at subsystem level components, which through the effective implementations of interfaces are interchangeable and upgradeable. Miller et al. (1995) address the issue of innovation for complex products. The signicance of this contribution notwithstanding, the study does not address the management of complexity for open or semi-open product architecture such as automobiles or heavy-duty trucks. Henceforth, this paperoriginally inspired by this previous research workaims at complementing the studies on technologic complexity by targeting the study of the complexity embedded in mass-produced modular product architectures such as automobiles. Hodbay (1999) claims that mass-produced commoditized products differ greatly from complex products and systems (CoPS) since these tend to be project-based, highcost and customized. Nonetheless, despite the fact that automobiles are mass-produced goods and many automotive product development and vehicle architectures could be considered commodities, some new automotive technologies show the development behaviour of CoPS, but at a modularized mass-produced scale. Here lies the rst gap in the available literature that this paper addresses: the absence of a study that tackles and describes the management of complex products for modularized and mass-produced products such as automobiles. Furthermore, several propositions on the evolution of modular technologic systems suggest that modular architectures evolve toward yielding more importance to subsystem modular developers than to product architects. Among the most relevant propositions affecting complex modular product architectures, we have selected the following ones: the propositions of technological leadership (Cusumano and Gawer, 2003); technological phases and modularity (Chesbrough and Kusunoki, 2001; Chesbrough, 2003); the necessity of a product integrator in complex products such as automobiles (Doran et al., 2007) and the technological independence of suppliers (Fixson and Sako, 2001; Olleros, 2007). This paper argues that the creation of technology complexity, especially for electric and electronic (E/E) automotive architectures seem to differ from the propositions of: (a) technological leadership (Cusumano and Gawer, 2003); (b) technological phases and modularity (Chesbrough and Kusunoki, 2001; Chesbrough, 2003); (c) the necessity of a product integrator in complex products such as automobiles (Doran et al., 2006) and (d) the technological independence of suppliers (Fixson and Sako, 2001; Olleros, 2007). The concept of modularity is the common foundation of all the above propositions. Among the reasons these propositions were selected are the following: to challenge

the applicability of Cusumano and Gawers (2003) proposition on technological platform leadership for industries for which product integrators compete (against other product integrators and specialized suppliers) on the base of a high-complexity product architecture, on a mass production scale and high liabilities like the automotive industry. Cusumanos proposition of technological phases (from modular to integrated architecture) is an interesting concept that may dene the relations among the actors of the automotive industry, especially where suppliers have become very specialized and have overtaken OEMs in their understanding of modular components. Propositions of Doran et al. (2007), Fixson and Sako (2001) and Olleros (2007) were also included since these authors question the necessity of a system integrator that dictates the pace of technological advancement. Their propositions assert that suppliers are themselves becoming, in their aim to increase their innovation capabilities, more independent from integrators. Therefore, the role of a product and technology integrator is being rendered obsolete. Moreover, Olleros (2007) claims that the actual involvement of systems integrators may hinder the innovative capabilities of suppliers. Fig. 1 shows three different automotive cases; the rst is bus and coach body manufacturing (low tech), the second is automotive interior design (e.g. seating systems, medium to high tech), and the third is E/E automotive architecture (advanced high tech). It also shows that the four propositions on the creation of technology complexity may be applicable to the rst two cases (bus and coach body manufacturing and automotive interior design). However, the applicability of those propositions to an advanced high tech case such as the automotive E/E architecture is highly questionable. The dotted arrows represent the questionability of the above proposals. For this research, the differences between the creation of technology complexityE/E automotive architectures and the above propositions have been encompassed in a single research proposal that aims to test them all against the innovation perspective of executives of advance engineering centres of OEMs, Tier-1 (T1) and Silicon Vendors, and specically about their perspective on drive-by-wire (DbW) systems. The relevance of this paper for non-automotive practitioners and scholars lies on the fact that brake-by-wire (BbW) and steer-by-wire (SbW) systems clearly convey the subject of study by confronting all of the above propositions with the reality of the complexity of technology creation. Furthermore, the study of DbW systems is relevant since many spin-outs of this technology can potentially impact other industries, such as consumer electronics and consumer silicon manufacturing. For automotive related scholars and practitioners; this paper depicts the managerial dynamics of innovation that take place within the automotive industry, especially for complex, innovative technologic developments such as DbW. Nonetheless, the dynamics described in this paper are similar and applicable to other automotive technologic developments, such as non-conventional powertrains, powertrain hybridization or automotive batteries development, among many others.

ARTICLE IN PRESS
C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485 475

Fig. 1. Advanced high tech cases such as the electric and electronic (E/E) automotive architecture have challenged the validity of propositions on the creation of technology complexity.

2. Research proposition The engineering of complex products has been an issue of study for practitioners and scholars, especially during the 1990s. Inuential works, such as Miller et al. (1995), have set the foundations for understanding the dynamics of value creation among the parties involved in the development of complex products. From this perspective, the constant evolution of automotive technology is leading toward the implementation of new, complex and expensive technologies in mass-produced vehicles. Hence, chassis electrication, hybridization of automotive drivetrains, and the inuences of other industries specically aeronautics and avionicsare making it possible to embed more complexity within an E/E vehicle architecture. This factthat product complexity is exponentially growing for an E/E automotive architectureand the industrys tendency to modularize subsystems are raising the assumption that the automotive industry is on the verge of a technologic and organizational disruption, where automotive OEMs are losing their competence and capabilities as technology developers to automotive suppliers (T1) and other lower-end vendors (T2), specically Silicon Vendors. Therefore, the following research questions arise as a result of the gaps in available literature:

Chesbrough and Kusunoki (2001) have argued that technology in modular architecture evolves in phases of openness and closeness depending on the technological stage of the product architecture. Cusumano and Gawer (2003), with their platform leadership proposition, fuelled up the assumption that rms tend to develop their own independent technological developments (comprised within intangible design boundaries). Olleros (2007) claims that non-contractual innovation (NCI) is better suited for enabling new independent modular technological developments since it stimulates design ingenuity and innovation; however, the author claims that contractual innovation does the opposite to suppliers. Particularly in the auto industry, Fixson and Sako (2001) claim that the specialized functionalities of automotive modular systems might bring independence of action and development to suppliers, making them product integrators. Also the ndings of Doran et al. (2007) claim that the industry is well advanced in the process of complete disintegration, in which T1 and other suppliers have started to increase their bargaining power and with OEMs becoming obsolete. Hence our research proposition, which has been designed to somehow encompass all of the above assertions, contends that: Increased complexity of automotive technology is causing OEMs to loose their core technological capabilities and their capacity to integrate within an automotive architecture that contains unknown technologies (to OEMs) and consequently switching the aim of innovation to very specialized automotive suppliers, which might ultimately lead the development of new technologies for automotive purposes. Automotive systems and in particular by-wire applications have emerged as a suitable test bed to investigate the adoption of new technologies in an environment characterized for having high production volumes and highly complex product architectures. Other industries/sectors

1. Is radical technological change in product subsystem modules causing a change in the relations within automotive networks? 2. Are radical technological changes in subsystem modular components modifying automotive networks in such a way that the automotive environment is changing from an OEM-centralized environment toward a more automotive peripheral subsystem environment?

ARTICLE IN PRESS
476 C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485

might be highly complex compared to automotive, for example aerospace/aircraft manufacturing is subject to high technological turbulence in domains that include materials science, avionics, turbine systems, etc. however the economics of production are different between the two sectors (airplanes are manufactured in the thousands, at most while motor vehicles are produced in the hundred of thousands). The work from Iasinti (1998) on technology integration is used to add and justify the element of integrating competencies. By-wire systems for automotive purposes are highly complex modular architectures which require several technical competences and the participation of several parties, all of them are eager to further advance the technology due to high potential returns. Thus the integration competencies of automotive integrators are eroded, due to the fact that they do not posses the technical, scientic or economical capabilities to integrate on their own (without the implicit participation of highly competent suppliers) newly developed technologies. 3. Research methodology The main purpose of this research is to contribute to the uncovering of the mechanics of technological adoption and value-creation interrelations among the parties involved in advanced engineering activities and practices for complex and modular product architectures, particularly for automotive product architectures. To achieve this, by-wire automotive applications were selected, specically BbW and SbW since those systems still have a great number of loose ends, making them very appealing from a research standpoint; they can illustrate the mechanics of adoption of new technologies in modular systems that are integrated within a larger technological system. For the purpose of this research, a comprehensive case study was constructed comprising the participation of automotive experts from different orientations and organizations, all related to the development of X-by-wire technologies. It is worth highlighting that the companies and persons selected for this study are actively involved in the understanding and development of by-wire systems for automotive purposes. This research includes the participation of:

The one- to two-hour interviews focused on the development of hi-tech automotive applications (by-wire and vehicular networks (VN)). The interviews, which were recorded, transcribed and coded followed the guidelines from Gerring (2007) and Yin (1994), and included key issues for this research work including: state of the art of available technology for braking and steering, value creation through innovation in by-wire activities, the role of suppliers in the technological evolution of braking and steering, the involvement of suppliers in the development of new automotive technologies, and knowledge creation and diffusion among the actors of automotive innovative networks. The justication of the methodology followed is based in the fact that innovative dynamics of highly turbulent technological developments can take place under consortiums (cooperative regimes between the involved parties). Consortiums are integrated by several layers of membership but core members (only a handful of companies comprising OEMs and T1 suppliers) are responsible for steering the direction of it. The participation of core members (although small in size number) is important in order to gain an insight into the dynamics of value creation and innovation for highly complex modular systems. References to similar approaches using a small number of participants include the work of Van den Hoed (2004) who produced a study on the use of fuel cell powertrains in the automotive industry. The ndings of this research work are presented using a structure that comprises an introduction to bywire systems, safety considerations for by-wire systems, architectural considerations for vehicles with by-wire systems, communication protocols for by-wire applications and, nally, a section where the results are discussed. 4. Technological overview of automotive by-wire systems The ndings section starts with a historical and technological overview of the status of available by-wire technology, followed by a detailed description of communication protocols and modular architecture for by-wire applications. This is necessary to understand the high complexity of the software interface between modular components. Finally, an expert assessment section is dedicated to discussing the adoption of this technology ` vis-a-vis the interaction between the different parties involved (OEMs, T1 and Silicon Vendors). 4.1. Origins and denition of by-wire systems Typically, by-wire systems refer to those systems that have substituted mechanical and hydraulic components with mechatronic actuators and on-board electronic control units (ECU) or dedicated computers for aeronautic or automotive purposes. Nowadays, by-wire systems are in full utilization in the aeronautical industry, where modern aircrafts have replaced mechanic or hydraulic ight control systems with full by-wire applications. The

 three advanced engineering directors from two different OEMs. One is Director of Advanced Engine Development (USA), one is Director of the Strategic Technology Development Ofce (Europe), and the other is Director of Advanced Vehicle dynamics (Europe); one Director of Engineering at a key T1 supplier of bywire applications: software, actuators and control modules (USA/Europe); one automotive innovation consultant and retired automotive OEM Corporate Ofcer (USA); two Senior Managers of automotive applications from a major Silicon Vendor (Europe and Japan) and three Research Scientists and Directors from automotive research centres in North America (USA).

   

ARTICLE IN PRESS
C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485 477

rst full by-wire aircraft was the NASA F-8C Crusader in 1972 (www.nasa.gov). Since the initiation of by-wire systems, automotive technologists have been trying to fully introduce this technology in all major vehicle systems. According to Kelling and Leteinturier (2003), by-wire systems can be classied in powertrain applications that include: (a) Gas-by-wire: whose main functionality is to provide the engine with the best possible mixture of gas and air; hence, throttle-by-wire is the most common system used in by-wire automotive applications. (b) Power-by-wire: encompassing a series of electronic peripherals aimed to reduced emissions. The most common application of power-by-wire systems is engine temperature management, after treatment and ignition-by-wire. (c) Shift-by-wire: whose main benet is to provide an adaptive transmission ratio to reduce the emission of CO2 and enhance vehicle performance and response comprising hydraulic automated transmissions, electro-mechanical transmissions, dual clutch transmissions and continuous variable transmissions. Kelling and Leteinturier (2003) identify that other by-wire applications could be implemented in the segment of vehicle dynamics; nevertheless, since these functionalities are tightly coupled with safety considerations, their full implementation has been slow. The most important applications that control the vehicle dynamics are: (a) BbW: which encompasses the transitory technology toward full electric brakes. BbW systems comprise electro-hydraulic braking (EHB) systems, electromechanical braking (EMB) systems and full electric braking (FEM). Among the most important efforts to commercialize BbW as an EHB system was the sensotronic braking system developed by Bosh GmbH

for Daimler-Chrysler in 2003, with its unfortunate well-known results. (b) SbW: which, without mechanical back-ups, comprises an electronic fault-tolerant system that enables the replication of mechanical handling of the vehicle and, at the same time, equals or exceeds the safety levels of current electric power steering systems. (c) Suspension-by-wire: refers to the electronic controllers and mechatronic actuators designed to increase vehicle stability and maneuvrability with no mechanical redundancy. The path toward full electric suspension is likely to be in stages that combine mechanical redundancies with mechatronic actuators and controllers. Currently, the most advanced semi-active damping systems can be found in premier vehicles such as Mercedes SL, BMW 7 series and VW Phaeton (Kelling and Leteinturier, 2003; Ayoubi et al., 2004). Fig. 2 depicts automotive by-wire systems already in use, partially adopted and still in development. The aim to implement BbW and SbW technology on vehicle dynamic applications continues; however, there are many considerations that need to be addressed on the current state of by-wire technology because it does not cope with the expectations of users and, on safety and economic bases, still does not comply with the requirements of OEMs. 4.2. Communication protocols for automotive purposes The automotive electronic architecture is constantly evolving; current automotive architectures have more than 80 ECU that are required to allow communication among them. The most important efforts of automotive OEMs and suppliers to enhance communication effectiveness within vehicle components include the development of a serial communication protocol called controller area network (CAN). This protocol was designed for transmitting

Fig. 2. Description of by-wire systems and their degree of development, according to Kelling and Leteinturier (2003).

ARTICLE IN PRESS
478 C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485

short control messages in real time (www.ni.com) and has been in use not only for automotive purposes but also for marine applications, agriculture equipment and heavy equipment. One of the most important features of CAN is its producer/consumer-oriented data transmission and multi-master capability (www.ni.com). As well, CAN has been standardized by the ISO 11898 for high- and low-speed/fault-tolerant versions. The CAN protocol for automotive applications such as engine control has a high speed of 1 Mbps; for automotive non-critical applications, such as body features or comfort devices, the data transmission speed reaches 125 Kbps (www.commons. scintech.susx.ac.uk). Another automotive complementary bus protocol, which is frequently used for automotive purposes, is the local interconnect network (LIN) bus. The specication of this protocol allows the communication of sensors and actuators operating at 12 V. This bus is mainly used for vehicle body applications such as sunroof actuators, intelligent wipers, HVAC and other body electronics applications. The LIN bus has a maximum speed of 1.92 Kbps over a maximum cable length of 40 m (www. interfacebus.com). Most of the time, the LIN bus is used as a subsystem within a CAN vehicle architecture (www. ni.com; www.interfacebus.com).

Fig. 3. Requirement comparison among communication protocols (modied from www.ni.com).

on a real-time basis (Kopetz, 2001; www.vmars.tuwien. ac.at). Under current technological conditions, the FlexRay protocol has clearly emerged as an industry standard for automotive by-wire applications; studying the rise of FlexRay as an industry standard goes beyond the scope of this paper. 4.4. Safety considerations of BbW and SbW

4.3. FlexRay and TTP fault-tolerant, real-time communication bus According to its developers, the FlexRay Consortium, the FlexRay protocol is a communication network that is a fault-tolerant, high-speed bus system (www.exray.com; www.ni.com). FlexRay provides error containment and delivers time determinism performance needed for X-bywire critical applications (www.comms.scitech.susx. ac.uk). Fault tolerance is achieved by allowing single- or dual-channel redundancy communication; hence, critical safety applications have the required critical redundancy, since data is transmitted using two channels. Error containment is achieved by an independent bus guardian, a mechanism that protects a channel from interference from faulty communication (www.ni.com). Typically, FlexRays data transfer rate is 10 Mbps on two channels or a gross data rate up to 20 Mbps. Fig. 3 depicts communication protocols for automotive purposes using the dimensions of requirements, data transfer rate and complexity. In this diagram, FlexRay and TTP/C are associated with a high degree of requirements, high data transfer and high complexity. In contrast, the communication protocols associated with LIN and CAN reect lower levels of data transfer rate, lower requirements and lower complexity. Time-Triggered Protocol/Critical system (TTP/C) is a timetriggered communication protocol for safety-critical systems that is appropriate for vehicle dynamics using X-byWire applications on a distributed real-time control system. This protocol is also suitable for non-automotive applications such as aircraft control systems, industrial power plants, or air trafc control. The TTP/C is a communication protocol that offers fault-tolerance features

Historically, safety considerations have been issues of the utmost importance for automotive manufacturers and suppliers. Regarding by-wire systems, some of the most important drawbacks of the state of the available technology comprise reliability and safety concerns. As previously discussed, the functionalities of by-wire systems are based on the communications and interaction between several ECU and other electro-mechanical actuators and the feedback that the system delivers to the vehicle operator. Hence, according to the ndings of the case studies, communications protocols must be dependable in several dimensions. Fig. 3 shows the dependability considerations for by-wire systems. Some of the most important issues that we believe inuence the overall adoption of by-wire systems are: dependable actuation, dependable computation, and dependable sensing (Kelling and Leteinturier, 2003); dependable power supply; and dependable communication. Among the most important issues for successfully implementing by-wire applications is the development of reliable, accurate, self-redundant and fault-tolerant communication software protocols and the related equipment needed to maintain effective communications (sensors, power supply, microprocessors, etc.). This guideline is based on IEC61508 and ISO26262 standards for programmable electronic safety related systems (Erkkinen and Conrad, 2007). 4.5. Steer-by-wire Power-assisted steering is not new to OEMs, Automotive suppliers or even drivers; nonetheless, this electric or electronic system still has the mechanical redundancy that ultimately provides minimum functionalities in case

ARTICLE IN PRESS
C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485 479

of system failure. Normally, SbW systems use multiple sensors, control algorithms and a reliable communication protocol to accurately transfer the steering input from the driver to the steering actuator (Brown and Holweg, 2004). Thus, according to Kelling and Leteinturier (2003), this subsystem (SbW without a mechanical interface) is the most unlikely among automotive subsystems to be introduced in the short to medium run. Some of the critical functions that need improvement are (Brown and Holweg, 2004; Kelling and Leteinturier, 2003; Ayoubi et al., 2004; and www.itz.tu-clausthal.de):

 Sensingspecically, sensors in the rotation angle and


torque for the steering device (wheel) and front tires must increase reliability (self-diagnosis) while reducing space (size) and power consumption. Actuationspecically, actuators for steering purposes must reduce power consumption, decrease the levels of noise and other acoustic considerations, increase ease of integration within the vehicle electronic architecture, be modular with interchangeable designs, and be able to be mass produced. Power electronicshigher power is needed for improved sensing and actuation; this scheme implies changes in the electronic architecture for at least 42 V power output for non-hybrids vehicles. Computationimproved and more reliable microcontrollers, especially with embedded FlexRay-ready compliance software, and suitability for automotive extreme operation conditions.
Fig. 4. BbW general depiction (Courtesy SKF. com, reproduced with permission of the copy right holder).

4.7. Expert assessment: case studies on the mechanics of technological adoption and value creation for by-wire automotive applications A questionnaire instrument was designed based in the elements of the communications protocols for automotive purposes and in particular the FlexRay protocol and bywire applications presented in previous sections. A sample of the questionnaire instrument submitted to interviewees is presented in Appendix A. Before applying the questionnaire instrument to the participants in this study questions were checked for the purpose of clarity and conciseness as this is a descriptive study. It is important to highlight that the interviewees have sufcient expertise to judge automotive architectural development, from its conception and in their evolutionary paths. Also the interviewees are FlexRay consortia members (core members and premium) and their answers were considered for this study. 4.7.1. Innovative multi-rm technological developments; the FlexRay standard The FlexRay protocol standard was originally developed by the BMW group as a ByteFligth (1999) passive protocol (for airbag release and other short time release applications). Daimler-Chrysler joined BMW to improve and further develop the ByteFlight for X-by-wire applications, which ultimately became the FlexRay protocol. The FlexRay consortium integrates several OEMs, T1 suppliers and Silicon Vendors. The main objective of the FlexRay consortium is to develop, with the collaboration of all its members, the establishment of the FlexRay protocol as a communication standard for on-board vehicle communication. The accomplishment of the case studies has revealed that, to achieve this, intensive collaboration among the parties involved is required. Specically, the two senior managers of automotive applications from a major Silicon Vendor and the automotive innovation consultant participating in the case study agreed that innovation depends on strong cooperation among parties. One of them claimed that, for automotive electronic purposes: Automotive innovation is only attainable by the strong cooperation of OEMs, Tier 1 and other suppliersy

4.6. Brake-by-wire BbW technology is designed to offer increased safety and vehicle stability to consumers and is expected to provide benets to automotive vehicle manufacturers, who will be able to combine vehicle components into modular assemblies using cost-effective manufacturing processes (www.ml.wpafb.af.mi). For end users, some of the expected yields of BbW are: shorter braking distance; optimized pedal feeling; limited vibrations; improved handling and overall driving safety (www.aa1car.com). According to Kelling and Leteinturier (2003), vehicle braking functionalities are in constant evolution toward fully electric braking systems, without mechanic or hydraulic redundancy. These systems are expected to hit the market in the short run. Among the considerations pushing for the implementation of BbW systems are: the increase of trafc density (which demands higher braking performance), user driving habits (oscillating speeds), environmental issues (high toxicity of braking uid), fuel efciency, and improved driving experience. Currently, BbW technology is heading toward a technological hybridization (Coronado et al., 2006) in which conventional braking systems are combined with EHB or EMB systems; Fig. 4 depicts how conventional systems act as a redundant application system for the electric part of the system. Fig. 4 shows the particularities of BbW systems.

ARTICLE IN PRESS
480 C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485

In light of the mechanics of technological adoption and value-creation interrelations investigated in this research work, for automotive (E/E) purposes (and for other complex subsystems), the most important opportunity for innovation and value creation still lies in the specialization of each individual activity performed by each member of a given innovative network. The same executive pointed out that specialization is the most important asset that a rm, such as T1 or Silicon Vendor, may have, because the criticality of the automotive electronic architecture may restrain architects to change to other suppliers since these products are unlikely to become commodities. Concerning a more proactive role of suppliers in electronic products architectures, and having more responsibilities for integrating subsystems by T1 and other suppliers, the Director of Engineering at a key T1 supplier of by-wire applicationssoftware, actuators and control modulesparticipating in the study (USA/Europe) expressed the following: We all have very clear and dened functions, so it is very unlikely that T1s and other suppliers are trying to encourage changes in their roles, especially if these do not concern themy A Senior Manager of a Silicon Vendor expressed the following: If OEMs seem not to understand the electronic architecture and are losing control of it, we will partner with T1 and Silicon Vendors to enable the OEMs to recover from this. It is actually bad for T1 or Silicon suppliers if OEMs ever lose control of the architecture; that means that they (T1 and Silicon Vendors) will have to take over the task of integration and they dont have an interest in doing thaty Hence, the risk of litigation on product faultiness and other judicial considerations play a key role in stopping the migration of T1 and other automotive suppliers to higher roles, especially to system integration. In reference to the point about the unwillingness to move toward integrators, the research scientists and directors from automotive research centres in North America (USA) agreed that potential litigations have stopped suppliers from becoming system integrators. One of them expressed the following: Liabilities and potential litigations risks are the main deterrents for automotive suppliers from becoming OEMs (or system integrators)y Thus, a complex automotive technology such as the FlexRay communication protocol for by-wire systems has to be developed in conjunction with all the actors of the innovative network, in this case the FlexRay consortium. Also, the research work shows that technologic competencies within automotive innovative networks are very specic; consequently, rms participating in this technological development (OEMs, T1 and other suppliers) tend to master their own specic area of knowledge while trying to avoid invading other rms areas of

competence. The research also shows that Cusumano and Gawers (2003) proposition of technological platform leadership, in this case the exercise by OEMs regarding the electric vehicle architecture, lacks applicability for this technological domain since technologic developments such as the FlexRay protocol (which has so many implications on the overall product) have to be developed and controlled by a large number of actors. Under this scheme, technologic advancements take place when individual developments done by a member of the consortium are brought in to be analysed and voted on, and, if accepted, implemented in the protocol standard. The complexity of technologic development such as the FlexRay protocol has highlighted the complexity of automotive subsystems. Fig. 5 relates high technological complexity with closed system architectures. The FlexRay protocol is located in the right upper corner of the diagram. Associated attributes of FlexRay include a high degree of value creation for OEM with a complex supply chain. That is in contrast to subsystems with low technological complexity (e.g. mechanical systems using standard components) with their associated low degree of value creation activities for OEMs.

4.8. Plug-and-play innovative developments; the AUTOSAR initiative The automotive open system architecture (AUTOSAR) is an industry initiative from OEMs and T1s that has the purpose of supporting the adoption of open standards for automotive E/E architectures, an indispensable feature for innovation in future automotive developments (www. autosar.org). According to AUTOSAR, some benets of the architecture are: scalability of vehicle platforms, integration of functional modules from multiple suppliers, and implementation of basic system functions commonality (www.autosar.org). Fig. 6 depicts the architectural openness of key competences in the automotive industry. Among the competences most likely to be locked by OEMs are chassis management, powertrain electronics and safety management (active passive). On the other hand, competences with no restrictions from OEMs include multimedia telematics, body comfort and humanmachine interface (HMI). Architectural openness has brought to other industriesspecically, the computer industrya ux of rms interested in obtaining the benets of introducing products that are suitable for the general architectural design without being committed nancially or strategically to the rms that developed such architecture (Garud et al., 2003). Christensen (1997) clearly describes how computer components (such as hard disc drives/HDD) rapidly evolved and ourished due to the establishment of an open and standard architecture. Nevertheless, this does not seem to be the case for all automotive E/E applications; the research work undertaken suggests that, for certain core competencies and core critical applications, OEMs are not willing to relinquish control to T1 or other components suppliers, specically due to the potential implications to human life in case of failure.

ARTICLE IN PRESS
C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485 481

Fig. 5. Complexity of automotive subsystems.

Fig. 6. AUTOSAR architectural openness (taken, enhanced and modied from: Fuelling the Future, Research and Advanced Engineering Innovation and Technology Report, Ford Motor Company, 2007).

The research work suggests that all technologic development in the areas of E/E such as chassis management, powertrain electronics and safety management systems are to be developed primarily by OEMs or in joint development between OEMs and T1 or other suppliers. On the other hand, the largest potential for plug-and-play modular applications lies in the comfort body electronics,

telematic applications and vehicledriver interfaces applications that do not interfere with critical electronic functions. The three Research Scientists and directors from automotive research centres in North America agreed that OEMs have to keep close control of the core capabilities characterizing open architectures. Moreover, they agreed that liabilities associated with product failure

ARTICLE IN PRESS
482 C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485

determine the openness of architectures. The opinion of one of these Research Scientists on open architecture is: Open architecture is good because it further opens the possibilities for more suppliers and partial innovation in top layers of the architecture, as long [OEMs] keep the core [capabilities] in close controly and Openness of architecture is determined by the liabilities associated with product failurey With regard to competence, the research undertaken also shows that OEMs are striving to update and further develop their competences in the following areas: engine development, safety critical systems; body engineering and total vehicle integration. 4.9. Additional implications of higher complexity of automotive technology: the case of VN The development of VN has witnessed an increase in the complexity of the technology employed. However, the path of technology complexity for vehicular networks (VN) may be similar to the path faced by internal communication protocols such as FlexRay. In Fig. 5, VN, as part of multimedia/infotainment/telematics, have competences with no restrictions from OEMs, as long as these applications do not disrupt critical vehicle functionalities (see Fig. 6). The potential benets associated with VN, more specically in the form of vehicle safety applications, comprise the propagation of collision-avoidance warnings and the exchange of driving parameters among vehicles for a safer driving. These messages primarily provide information about current road conditions or trafc status in the vicinity. Given the nature of safety applications, the exchange of information needs to be distributed on a real-time basis since they are labelled as critical or life threatening. Hence, this technology (VN) might have to face similar challenges as internal communication protocols (such as the FlexRay protocol, i.e. high-robustness, fault-tolerance and self-redundant), especially for the transmission and exchange of critical safety information such as collision warnings and other life-threatening situations. In general, these applications comprise vehicle-to-vehicle (V2V) communications, though another type of vehicular application considers vehicle-to-infrastructure (V2I) communications. The latter is mainly intended for non-safety applications, such as the case of retrieving information from the xed network to the vehicular environment. According to the specication given to the vehicleinfrastructure integration (VII) report based on direct short range communication (DSRC) technology , the following elements are briey described:

the interaction between the customer and information services. The OBE can enable communication links with the vehicles in the surroundings as well as with the roadside infrastructure. Roadside Equipment (RSE)This element is considered the communication interface or gateway between the xed infrastructure and vehicles. It comprises a DSRC transceiver (RSU), a GPS location system, an application processor, and a router to connect to the xed network. To communicate with the vehicles, the RSU transmits a list of services contained in a provider service table (PST). Any raw information coming from an OBE is rst veried in the RSU processor and then forwarded by the router to reach the xed network.

Information services in vehicular environments represent promising business opportunities, especially on V2I, where the supply of information services can be offered by heterogeneous service providers dwelling in the xed network. The successful diffusion on a mass scale of by-wire systems is the cornerstone for the implementation of even more complex technologies such as active chassis control (ACC) and operational VN. Therefore, the effective applicability of ACC will, for example, enable management of the safety and security warnings (rollover occurrence, frontal, lateral or rear collisions, etc.) of a vehicle. Consequently, ACC in combination with VN will allow transmission via OBE to other vehicles within the transmission range (to warn drivers to take necessary precautions) and, in case of an urgent warning, to transmit emergency response services via RSE. Despite the early stages of this technology (VN), it is possible to foresee a similar technological creation process for which a technological consortium (which encompasses OEMs, T1, lower level suppliers, academic research centres and universities) ought to set the pace and evolution of these technological systems (VN).

5. Conclusion The research presented in this paper shows a two-fold phenomenon taking place in the automotive industry. First, it is very unlikely that Automotive OEMs are losing control of modern automotive technologies such as E/E architecture functionalities and associated BbW systems. Automotive OEMs, contrary to what has been proposed by Cusumano and Gawer (2003), do not exercise a technological leadership on E/E issues, due to the immense complexity of developing these technological systems. What OEMs do is develop loose collaboration agreements among all members involved in the innovative network, which enables a transparent and unhindered ow of information to all parties involved in these technological developments. It is actually better for OEMs, T1 and Silicon Vendors to make their individual research outcomes available so they can spread the associated research development costs among many participants. Furthermore, bodies such as the FlexRay consortium or

 On-board equipment (OBE)This element is installed in


the vehicle and comprises an on-board unit (OBU) that is a 5.9 GHz DSRC transceiver, a GPS location system, a processor for application services, and a HMI to handle

ARTICLE IN PRESS
C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485 483

the AUTOSAR initiative enable their participants to maintain the technological pace of new developments by promoting the active participation of their members in the evolutionary path of these technologies. Second, the research presented in this papers shows that it is true that automotive suppliers (T1 and Silicon Vendors) are becoming more and more specialized. This situation, plus the higher requirements set by electried powertrains, causes the technological complexity of the E/E vehicle architecture to constantly rise, thus making it more difcult for all parties to fully understand the complexity embedded in modern E/E architectures. Consequently, the only way to make further advances to the technology is to have the active involvement of all parties concerned (OEMs, T1, Silicon Vendors, etc.) in the understanding of the technology. It is actually negative for T1 and Silicon Vendors to overtake and set aside OEMs in the development of new technologies since this would imply that suppliers have to perform the integration tasks that OEMs normally would do and they would have to bear the liabilities and litigation risk that OEMs bear. This paper argues that the propositions of Chesbrough and Kusunoki (2001) and Chesbrough (2003) of technological phases, despite their claims of the automotive industry being technologically vertically integrated, seem to be inappropriate for automotive modular technologic developments since these developments evolve horizontally in conjunction with all the members of the innovative network. In relation to the claim of Doran et al. (2007) that automotive developments can override OEM integration processes and control, the research presented shows that, while this may be true for low-tech, commoditized standard components (such as conventional components) and low-end vehicles, their proposition fails to acknowledge that OEMs play the important role of liability bearer, and this fact is a key entry barrier for Automotive suppliers to become OEMs. Their claim lacks applicability to complex technological developments with a high degree of turbulence, such as BbW systems. This kind of development requires the committed participation of all members of a technological network, as well as specialized rms that master their domain of expertise and are willing to share their information with the other members of the development network. In relation to Fixson and Sako (2001), automotive suppliers have indeed become a key component in the development of new technologies. However, this involvement does not translate into a carte blanche for total development independence from OEMs; technological developments that do not consider OEMs are less likely to be implemented on an actual vehicle architecture because OEMs are not willing to be bound to a technology they do not fully understand. Also Olleros (2007) proposition on NCI seems to lacks applicability for automotive critical systems and other complex critical safety applications, such as aeronautics applications (ight control systems or landing gears). These developments are created with the involvement and implicit consent of all parties affected by these developments,

especially OEMs. Nonetheless, his proposition might be applicable for non-critical applications, such as Infotainment and convenience electronic applications. The Research Scientists and Directors from automotive research centres in North America agreed that automotive suppliers should not develop standards on their own. As expressed by one these Research Scientists: It is very dangerous for automotive suppliers to attempt to develop standards without the implicit consent of OEMsy BbW systems are being held back by the actual conformance of the current vehicle architecture; nonetheless, powertrain hybridization and improvements to power electronics are pushing this technology forward. It is important to bear in mind that those BbW systems, on their own, represent a very poor sales argument for OEMs since mechanic or hydraulic systems might have similar or improved performance at much lower cost. The real importance of BbW systems is that they are enablers of higher and more complex technologies such as active chassis control (ACC)/unied chassis control (UCC) and intelligent vehicles (IV) (Bishop, 2005). It is possible to foresee that the integration of BbW and SbW technology into more complex vehicle dynamics systems such as ACC/UCCwhich, combined with external vehicular networks, radar and other sensorswill ultimately enable the vehicle itself to recognize life-threatening situations and undertake immediate actions (such as lane changing, emergency braking and deployment of collision mitigation countermeasures) in case of imminent danger to passenger safety. It is important to acknowledge that similar dynamics might take place in other industries such as construction equipment, marine applications, heavy-duty vehicles or medical equipment since they intensively use mechatronic actuators.

Appendix A A.1. Sample questionnaire submitted to interviewees 1. Please briey describe the state of the art of the available engine technology and compare it to the state of electric drives. 2. According to your understanding and knowledge, what are the main deterrents to the commercial uprising of electric power drives? What are the implications of electric power drives in relation to other vehicle components, such as by-wire systems? 3. In reference to the previous question, are technological limits set by the available E/E architecture? Describe the role that suppliers play in the development of electric and none conventional drives. Is it critical? 4. According to your experience and knowledge, is the technological advancement of vehicle architectures especially electronicssomehow limited by the state of the technology of other technological domains (e.g. battery range, feasibility of 42-V architecture, robustness and capacity of available silicon equipment)?

ARTICLE IN PRESS
484 C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485

A.2. Questions about innovation and value creation dynamics of the technology 5. According to your knowledge, are automotive suppliers becoming more specialized in certain technical domains, such as by-wire systems? Generally, do you think that the expertise in advanced electronics and communication software protocols (TTP and FLEXRAY) could ever be replicated by individual OEMs? To what extent would you describe your collaboration between OEMs and AS for the purpose of developing highly advanced vehicular systemslow, medium, and high? Explain why. 6. Describe the implications to your work (development of electric drives) of the increasing complexity of other vehicular systems, particularly DbW. Is it determinant for OEMs to keep the pace with these technological advancements despite it not being the core of their value creation activities?

A.4. General questions 12. Do you see a transition in the automotive industry in general toward OEMs yielding their importance to AS developers of critical and highly advanced subsystems? Do you see the technological complexity moving from being controlled by OEMs to being understood and enhanced by AS (e.g. DbW and vehicular networks)? 13. Do you believe that the risk of litigationcivil and criminal suitsinhibit the development of automotive innovations such as DbW (especially in safety critical systems)? 14. Is standardization the missing component for the technology (in electric drives and other applications such as DbW) to take off?

Appendix B B.1. Participant rms

A.3. Questions about the innovative environment of technology and relation within the industry 7. In your opinion, does the actual regime of collaboration between all parties of the automotive innovation chain (Silicon Vendors, T2, T1, and OEMs) enable the investment and efforts toward the advancement of the technology? Are they fair for all the members of the innovative network? 8. According to your understanding and knowledge, does the majority of technical innovation impacting automotive electronic architectures come from suppliers or OEMs? 9. Compared to 10 and 5 years ago, do you see broader and more decisive participation of AS in vehicle electronics architectural denitions? Do you foresee a more inuential participation of AS in the development of the necessary technologic and scientic advancements that will be needed for future vehicles (especially on electronics)? 10. Automobile and other motor vehicles as complex products encompass several scientic and technological domains, making it difcult to cope with the pace of technology for OEMs; new technologies such as DbW systems are being developed with the aim of having an impact on automotive architectures. According to your understanding, are these innovations (among others) developed with the strategic aim to give more scientic dominance to the automotive developers of these technologies? Do you foresee a trend toward integrating technological and other scientic capabilities within OEMs internal capabilities? 11. According to your perspective, knowledge and experience, is it worthwhile for OEMs to further open vehicle electronic architecture? How would you explain the increasing participation of supplier in infotainment applications and not in other more critical applications?

For condentiality reasons, the name of the participant rms cannot be disclosed for publication. References
Ayoubi, M., Demmeler, T., Lefer, H., Kohn, P., 2004. X-by-Wire Functionality, Performance and Infrastructure. Convergence Transportation Electronics Association, Detroit, MI. Baldwin, C.Y., Clark, K.B., 1997. Managing in the age of modularity. Harvard Business Review, SeptemberOctober, 8493. Bishop, R., 2005. Intelligent Vehicle Technology and Trends. Artech House, Boston, MA. Brown, S., Holweg, E., 2004. Smart Electro-Mechanical Actuation for Drive-By-Wire Applications. Convergence Transportation Electronics Association. Camuffo, A., 2000. Rolling out a World Car: Globalization, Outsourcing and Modularity in the Auto Industry, International Motor Vehicle Program. MIT, Cambridge, MA. Chesbrough, H., Kusunoki, K., 2001. Modularity trap: innovation technological phase-shifts and the resulting limits of virtual organization. In: Nonaki, I., Teece, D. (Eds.), Managing Industrial Knowledge: Creation, Transfer and Utilization. Sage, London, UK. Chesbrough, H., 2003. Towards a dynamics of modularity, a cyclical model of technical advance. In: Prencipe, A., Davies, A., Hobday, M. (Eds.), The Business of System Integration. Oxford University Press, Oxford, UK. Christensen, C., 1997. The Innovators Dilemma. Harper Business Essentials, New York, NY. Cusumano, M., Gawer, A., 2003. Platform Leadership: How Intel, Microsoft and Cisco Drive Industry Innovation. Harvard Business School Press, Boston, MA. Doran, D., Hill, A., Hwang, K., Jacob, G., 2007. Supply chain modularisation: cases from the French automotive industry. International Journal of Production Economics 106, 211. Erkkinen, T., Conrad, M., 2007. Safety-Critical Software Development Using Automatic Production Code Generation. MathWorks, Inc. Fixson, S., Sako, M., 2001. Modularity in Product Architecture: Will the Auto Industry Follow the Computer Industry? International Motor Vehicle Program. MIT, Natick, MA. Garud, R., Rattan, N., Langlois, R., 2003. Managing in the Modular Age: Architectures, Networks and Organizations. Blackwell Scientic Publications, Oxford, UK. Gerring, J., 2007. Case Study Research: Principles and Practices. Cambridge University Press, Cambridge, UK. Helper, S., MacDufe, J.P., Pil, F., Sako, M., Takeshi, A., Warburton, M., 1999. Project Report: Modularization and Outsourcing: Implications for the Future of Automotive Assembly, International Motor Vehicle Program Annual Forum. MIT, Cambridge, MA.

ARTICLE IN PRESS
C.C. Mondragon et al. / Int. J. Production Economics 118 (2009) 473485 485

Kelling, N., Leteinturier, P., 2003. X-by-wire: opportunities, challenges and trends. In: Vehicle Networks, Safety Critical Systems and Accelerated Testing and Reliability, SAE Technical Series. Kopetz, H., 2001. A Comparison of TTP/C and FlexRay. Institut fur Technische Informatik, Technische Universitat Wien, Austria. Langlois, R., Robertson, P., 1992. Networks and innovation in a modular system: lessons from the microcomputer and stereo components industries. Research Policy 21, 297313. Miller, R., Hobday, M., Demers, T.-H., Olleros, X., 1995. Innovation in complex systems industries: the case of the ight simulation. Industrial and Corporate Change 4. Olleros, X., 2007. The power of non-contractual innovation. International Journal of Innovation Management 11 (1), 93113. Yin, R., 1994. Case Study Research: Design and Methods. Sage, Beverly Hills, CA.

Web Sites
/www.aa1car.comS. /www.autosar.orgS. /www.commons.scintech.susx.ac.ukS. /www.interfacebus.comS. /www.itz.tu-clausthal.deS. /www.exray.comS. /www.ml.wpafb.af.miS. /www.ni.comS. /www.vmars.tuwien.ac.atS.

You might also like