Toward Software-Defined Vehicles: How a leading OEM leverages product architecture to reconfigure automotive software business ecosystems.

Publication Type:

Conference Paper

Source:

Gerpisa colloquium, Paris (2020)

Keywords:

ecosystems, OEM, Product Architecture

Abstract:

The automotive industry is currently facing one of the more radical transformation in its history through the rapid emergence of Connected, Autonomous, Shared and Electric (CASE) vehicles. These trends are leading to an unprecedented technology paradigm shift, from innovation trajectories centered on mechanic technologies to ones centered on digital technologies, mainly driven by industry newcomers. In order to face such a paradigm shift, OEMs must adapt their activities and reconfigure the structure of their business ecosystems, especially around software development issues.

Ecosystem structure can be thought of as a multilateral alignment of its members’ activities, balancing the optimization of both collective mechanisms for value creation and individual mechanisms for value capture (Jacobides et al., 2006; Adner, 2017). Ecosystem reconfiguration looks quite challenging since it implies to transit through an uncertain path from a well known and commonly agreed ecosystem structure to a new one that can threat some actors’ roles and even lead to their “death” (Moore, 1996; Adner, 2012). Many scholars in ecosystem literature point product modularity as a critical lever to structure business ecosystems and to support their members’ collective efficiency (Moore, 2006; Jacobides et al., 2018). This idea find its main support in the "mirroring hypothesis" brought by scholars working on product architectures and stating that the structure of organizations tends to mirror the architecture of the products they develop (Colfer et Baldwin, 2016). Following this hypothesis, product architectures - the scheme by which product functions are allocated to its components, including its functional elements, its components and their interfaces (Ulrich, 1995) – can be seen as templates that influence the structure of business ecosystems. Modularity is then pointed as a good design strategy to optimize both firm's boundaries and coordination mechanisms (Baldwin & Clark, 2000) while reducing their transaction costs (Langlois, 2003). Product architectures can thus appear as useful levers to reconfigure business ecosystems, especially in a context of technology paradigm shift, but are also known to be quite hard and long to transform because of organizational and cognitive resistance (Henderson & Clark, 1990). Surprisingly, our understanding of ecosystem reconfiguration only find very limited supports in the literature and our understanding of the role of product architecture in ecosystem reconfiguration suffers from an even greater gap, leading us to focus this contribution on the following research question : how a firm can leverage its product architecture to reconfigure its business ecosystem? To address this research question, our contribution builds on the case of one of the leading global OEM, which is currently involved in the reconfiguration of its ecosystem through radical transformations of its vehicles Electric and Electronic (E/E) architectures.

This study was carried out in collaboration with a leading OEM’s software engineers team gathering more than 400 software architects responsible for some of the key transformation in vehicles’ E/E architecture (E/E architecture includes both vehicles’ electronic components and software). These settings allowed us to get access to a wide variety of data on which we built an in-depth case study aiming at identifying how this OEM re-design its vehicles’ E/E architecture to manage ecosystem reconfiguration. The data collected include: 1) 29 semi-directed interviews with software architects and employees responsible for the software development strategy; 2) dozens of internal documents either related to the OEM software strategy or to software architecture (technical) or to project management and internal processes; 3) Field notes related to informal discussions or to working sessions with engineers; 4) external data collected from press reviews or industry experts reports and databases. First, these data allowed us to understand and represent the current ecosystem structure and technology trends around software issues. Second, we were able to analyze the evolution of the vehicle’s E/E architecture across 3 generations of architecture corresponding to vehicles: 1) released for the past 5 years; 2) that will be released for the next 5 years; and 3) that will be released from 2025 (in development). We then analyzed how these architectural changes are affecting the structure of this OEM ecosystem, focusing on software development activities.

Through this study we found that in order to face the increasing complexity of automotive software and the rapid expansion of vehicles’ features the OEM evolves its vehicles’ E/E architecture following a logic that we framed as “functional transposition”. This logic is supported by three key changes: the “abstraction” of software from hardware; the increased integration of the hardware architecture; and the implementation of a common modular software platform for the entire set of vehicle’s features, based on a “Service Oriented Architecture” (SOA). These three changes are expected to allow the OEM to solve technical bottlenecks related to vehicles’ data management but also to adopt a central position in vehicle software development by improving its ability to orchestrate complementors. Overall, we also point that this OEM explicitly take into account ecosystem structure as a critical criteria to make architectural choices. It thus consciously leverage product architecture to reconfigure its ecosystem.

First, this study contributes to theory by providing an empirical support enlightening the critical role of product architecture in the reconfiguration of business ecosystems. It thus contributes to the ecosystem literature which is suffering from a lack of study focusing on the reconfiguration of ecosystems but also from a lack of empirical supports exploring the links between product architectures and ecosystem structures. We also contribute to modularity theory by conceptualizing the logic of “functional transposition” which does not only relies on a variation of the degree of modularity but rather on the nature of modularity. Second, we contributes to practice by giving inspiring insights about how firms in the automotive sector can adapt and evolve their roles. Also, our insights could be useful beyond the automotive sector since many sectors are subject to disruptions coming from digital technologies. The logic of “functional transposition” might be a useful pattern to drive ecosystem reconfiguration in some of these sectors too.

GIS Gerpisa / gerpisa.org
61 avenue du Président Wilson - 94230 CACHAN
+33(0)1 47 40 59 50

Copyright© Gerpisa
Concéption Tommaso Pardi
Administration Géry Deffontaines

Powered by Drupal, an open source content management system
randomness