Automotive electronics: Bumpy road ahead
At the Automobil Elektronik Congress in Ludwigsburg near Stuttgart, those responsible in the automotive value chain for E/E architecture, vehicle connectivity, autonomous driving and electrification of the powertrain allowed a glimpse of the cards they are holding. Comprehensive reporting on all topics of the two-day event would go beyond the scope available, so eeNews Europe will confine itself to summarising what we consider to be the most important statements in keywords.
Business models: Disruption comes in sight
Most representatives of the automotive industry (not only in Germany) already suspect that the electrification and automation of driving will greatly change their business model. Wu Gansha, CEO of Chinese technology provider UISEE, gave an idea of what this means. The company aims to establish a nationwide mobility network in China by 2031, consisting of electric, self-driving vehicles that are designed for shared use. Wu believes that this fleet will not only eliminate traffic accidents and traffic jams, but also reduce the costs for mobility and logistics for the user to one third of the current level. With China being the world’s largest market for passenger cars, it can be expected that implementing such a business model would certainly affect the way carmakers are designing and people are using cars across the entire world. Wu compared the situation of the car industry today with that of the bicycle industry a few years ago: As recently as 2014, 79 million bicycles were produced in China. Following the introduction of shared-usage models, the number of bicycles produced fell by 33 percent in just three years. It is quite possible that a similar development is imminent for the automotive industry.
In any case, as the various presentations also made clear, the role of car manufacturers is changing from being vendors of a very complex and expensive consumer article to providers of mobility services. This applies to tier 1 suppliers as well as OEMs. Rolf Zöller, who is responsible for the entire electrical and electronics development in the Volkswagen Group, explained what is particularly important to the company: “We do not want to give up the interface to the customer,” said Zöller in his keynote address. Because the company wants to reach its customers directly via its services in the future as well, mobility services are to remain one of the core competencies in the hands of the car manufacturer, said Zöller.
The map supplier Here, a joint holding of Audi, BMW and Daimler to provide the real-time map material for autonomous driving, also wants to expand its business model: The company plans to make the data available not only for real-time navigation of vehicles, but also for indoor navigation in the future. Overall, says CEO Edzard Overbeck, Here will expand its business focus on 3D navigation in Smart Cities and offer navigation services for all vertical layers. “This (vision) is not 30 years away, it is just around the corner”, Overbeck said.
Technology: Lifeboat of the existing automotive industry
Of course the car industry is not inclined to surrender without a fight. Instead, the industry is trying to defend itself as well as possible despite imminent interruptive changes. And of course, technology is an essential element of this defense strategy. Thus, Volkswagen’s Zöller sketched the scenario of a comprehensive software environment for mobility. “Software orientation is becoming Volkswagen’s core competence,” said Zöller. The VW top developer announced the standardization of the software basis for the vehicles in order to facilitate software reuse and to create the possibility of easily integrating software modules with basic functions. This “Volkswagen OS” will be much more than a conventional operating system, Zöller highlighted. And, in the sense of reducing complexity, “standardization is more important than leading edge (technology),” explained Zöller.
One of the favourite children of today’s automotive software is Artificial Intelligence (AI). But on this field, wishful thinking often obscures the view of reality. In his presentation Gabriel Seifert from management consultancy Accenture showed where the gaps and technology construction sites of KI lie.
After that, AI applications requires a very high training effort. The technology is also inflexible and has difficulties in dealing with exceptional cases. Moreover, AI-controlled systems are very susceptible to manipulation; when interpreting images, adulterations can be incorporated at the bit level, which humans do not even notice, but which lead an AI to a completely different conclusion. Gabriel showed pictures of tampered stop signs that were supposed to cause AI-equipped vehicles to stop – in practice, however, there were cases where the vehicles instead accelerated in the face of the signs. And last but not least: The decision-making process of AI systems is not transparent and comprehensible and therefore difficult or impossible to reproduce. “AI is a black box,” Gabriel said. “There are massive possibilities for manipulation.” For all these reasons, many OEMs in automotive industry are currently reducing their AI-related development activities.
More articles on specific presentations and technology introductions will follow over the week ahead.