Is software the new frontier for the Automotive-turned-Mobility Sector?

Type de publication:

Conference Paper

Source:

Gerpisa colloquium, Detroit (2022)

Mots-clés:

design capacities, industry architecture, mobility industry, Software

Résumé:

The global automotive industry is deeply and firmly committed to a transition to carbon-free mobility. But electrification is proving to be the first step in an even deeper transformation of this industry. What most of the new entrants in the pure electric car sector, such as Tesla, NIO, and Xpeng, have in common, besides the fact that they only offer battery electric vehicles, is that they leverage an innovative software architecture that potentially allows them to outpace incumbents. When Foxconn recently presented “three self-developed electric vehicle models”, its promotion of the performance of these (still prototype) vehicles was less noteworthy than its goal of “…transforming itself into a software-driven enterprise for the future.”
Common to all of these examples is the ambition to start with the software (i.e. a vehicle operating system) and to build the hardware around it, i.e. aiming for Firmware Over The Air (FOTA), wide connectivity to almost everything, and autonomous driving to be native functions of these vehicles, with AI and machine learning being applied throughout. While incumbent OEMs have, so far, taken a function-by-function approach to developing software, these newcomers are attempting to turn the potential handicap of starting from scratch into a major advantage, namely a more open and scalable software and hardware architecture.
This software revolution, as it unfolds in the automotive industry, will have very significant impacts on design capabilities and industry organization. Will incumbents be able to develop the capabilities to compete in software-centric mobility as they have, so far, been able to do with electrification?
Connected, Automated, Shared and Electric Vehicles (CASE) have, in fact, become the new frontier for the automotive industry. Beyond the costly R&D required to develop this new generation of vehicles, the crucial challenge is that the automobile is no longer the top-level system but a component, albeit a major one, of the new mobility systems. This implies that the vehicle must be connected to its direct environment, i.e. the road infrastructure, but also other vehicles, pedestrians, the home, the power grid, etc., as well as the front and back ends of mobility services. In other words, the vehicle is deeply integrated into a mobility services architecture. Expanding on a presentation made by T.Fujimoto and JP.MacDuffie at "The new mobility" conference held in November 2018 in Tokyo, we build an analytical framework consisting of four layers: the low ground (i.e., the road infrastructure), the ground (the vehicle), the low sky - all the connections between the vehicle and its environment - and finally the high sky, where the mobility services are located.
This framework will help us characterize the new software capabilities that automakers, whether incumbents or new entrants, are developing to position themselves in software-centric mobility. Software is not new to the automobile or the automotive industry; software standards for in-vehicle functions have long been common, such as CAN (Controller Area Network), developed by Bosch in the 1980s. Engineers with mechanical expertise have learned to write the necessary code or outside IT contractors are used, leading to the function-by-function approach.
For the coming era of software-first, FOTA-enabled designs that fully integrate in-vehicle functions with external mobility services, incumbents face both expanded scope and a need for architectural (i.e., relationships among components) innovations that are challenging their current capabilities. Do they stay in their safe zone, i.e., the vehicle, or do they develop more layers of the software stack at the heart of the mobility system architecture? How can they tackle the more ambitious goals? Do they hire thousands of highly skilled programmers or do they make strategic alliances or do they buy specialized companies? When competing for software talent with other industries, can they be successful in attracting and retaining that talent?
We collected data in three ways. First, we provide a descriptive case study based on observation of the integration of carsharing enablers into an existing electric vehicle whose software was developed using a function-by-function approach to document how this is accomplished. Second, to comprehend the strategic thinking of automakers, we examined public statements of their senior management teams about mobility services and software development – and then probed further via semi-structured interviews with top executives of a few automakers. Third, we gathered public information from automaker and IS/IT company websites on mobility services strategies and organizational arrangements.
We observe that all the technology giants have already claimed a place in the automotive industry: Stellantis has announced a major partnership with Amazon, Ford and Renault are already integrating some of the main functions proposed by Google and Microsoft has undertaken some projects about autonomous driving with VW and GM. Finally, Alphabet has been investing in autonomous mobility services for nearly 15 years through its subsidiary Waymo, while rumors about an autonomous car developed by Apple have been going on for nearly 10 years.
Similarly, electric newcomers, Tesla and Chinese New Energy Vehicle manufacturers such as NIO and Xpeng, to name some of the best known, are all offering vehicles with software at the core of the design. They are built around a vehicle operating system that enables high connectivity with the immediate environment, scalable autonomous driving capabilities, and provides an open platform for implementing services, both mobility and non-mobility, such as the NIO App that provides access to many third-party services.
In this rapidly changing environment, driven by companies from outside the traditional automotive industry, incumbent automakers, burdened by the weight of past investments in their current product lines and the need to maintain their conventional combustion vehicles, are finding it difficult to implement disruptive strategies to catch up with their new competitors, whether they are technology companies or highly dynamic new entrants. Indeed, incumbent automakers have to make a technological leap to enter the new mobility services market, for which they must also take on the challenges and risks. Vertical integration may prove more attractive – or more necessary – than past approaches that have relied heavily on outsourcing to IT suppliers.
From a practical point of view, in the short term, automakers could lose ground to new competitors, not because of the poor performance of electrified vehicles, but because of products that are less open and less easily adaptable to the requirements of new mobility services or, more simply, more complex to use because they are too far from the standards of technological products. In the longer term, the inability to enter the new world of mobility services, due to the incapacity to transform into a software driven company, could jeopardize the industry's sustainable business model of selling products to individuals or fleets. Technology firms wanting to enter the automotive industry face their own challenges to master many new capabilities necessary for designing, manufacturing, and coordinating the supply chain for such a large complex multi-technology product. Which challenge — tech learning automotive hardware or autos learning how to master software — is greater and who will learn faster?

Video

      

 

  
GIS Gerpisa / gerpisa.org
4 Avenue des Sciences, 91190 Gif-sur-Yvette

Copyright© Gerpisa
Concéption Tommaso Pardi
Administration Juan Sebastian Carbonell, Lorenza MonacoGéry Deffontaines

Créé avec l'aide de Drupal, un système de gestion de contenu "opensource"
randomness