ICT Supports Digital Transformation to Smart Cars
This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>
Enterprise products, solutions & services
Making cars smart enough to drive themselves takes a lot of technology, and the ICT industry can look forward to big opportunities in providing it. From processors to an infrastructure for the Internet of Things (IoT), ICT firms will need to supply the technology that companies such as FAW Group Corporation (FAW) require to bring about the digital transformation of driving.
Today, the automotive world is accelerating the development of truly smart cars. However, digitalized smart car technology is still rather new and immature, and we need to ask, ‘What does it mean for the future of the automotive industry?’ As China’s first automotive enterprise to conduct R&D on smart cars, FAW recognizes that we must design our own road to digital transformation.
What is a smart car? The Society of Automotive Engineers (SAE) in the United States offers the most widely accepted definitions. The immediate goal for the majority of automakers is to reach SAE’s Level 3 for smart cars, or Conditional Automation, where automobiles control speed and steering programmatically and rely on the human driver to take over in dynamic situations, like when bad weather interferes with the car’s sensors. Level 3 cars may become widely available by 2020. To get there, smart cars will need more than their own real-time monitoring and braking capabilities; they will need Artificial Intelligence (AI), Big Data, cloud computing, and other ICT technologies to achieve the deep integration of smart hardware and software for the necessary levels of safety and autonomy.
While driving can be a great pleasure, it is also troublesome. For example, in the United States, deaths caused by traffic accidents average about 20,000 a year and are not declining. The U.S. Department of Transportation believes that the collision-avoidance aspects of smart cars will help solve this difficult problem — and so the government is encouraging high-tech enterprises such as Google to use their technology to mobilize smart car development. The goal is to position the United States as a leader in the smart car industry, along with the European Union and Japan. A strategic goal in Japan is to lead the world in automated driving/Vehicle-to-Everything (V2X) standards.
FAW must choose the technologies that will be developed in-house versus acquiring others that will enable the transition from cars to ‘smart’ cars. By equipping cars with more sensors, processors, and software, passenger vehicles will become integrated carriers of digital transformation. FAW understands that an increasing amount of the technology for intelligent systems will be in the cloud, especially as it is required to work in the Big Data environment. Here is FAW’s summary for smart car digitalization:
Car + IoT — The result of ‘Car + IoT’ and ‘Car + Internet’ will expand the scope functions that cars are equipped to handle, with the catch that this expanded scope also demands a transformation of the car companies themselves. Traditional, manufacturing-dominant enterprises must adapt to become service-oriented. Quite simply, future automotive enterprises must deliver customer benefits through services, or they will not succeed.
Car + AI — Three important domains require integration with AI in order to create smart cars. The first is sensor fusion, the second is route planning, and the third is using AI and Big Data for multiple levels of data classification and delivery of results. Currently, China is working on AI 2.0 under the guidance of the Chinese Academy of Engineering. In the immediate future, smart cars will mainly use on-board AI capabilities, with supplementary support by cloud-based AI services. As cloud and ICT technologies develop further, cloud-based AI support will become the primary director of smart cars. For example, cars will need a basic level of onboard intelligence. They must be able to determine if objects in the environment are people, cars, or a barrier. On the road, cars must determine the speed and direction at which cars and people are moving. To do this, the widest possible spectrum of data must be integrated and available to the AI support system.
Car + smart manufacturing — For smart manufacturing, cars are the most important and most promising industry. Traditional car manufacturing is serial: Starting with product planning and engineering design, then to experimentation and trial production, and from full-scale production to marketing and post-sales services. The emergence of enterprise cloud capabilities from advanced ICT infrastructures is pushing the entire car production effort into many concurrent processes using virtual platforms. Major components within the automotive enterprise cloud include digital design, manufacturing, and service platforms, that together greatly improve the efficiency and cost effectiveness of the entire organization. For example, when the Hongqi car was being developed for the Chinese market in the 1950s, achieving Europe’s 5-star collision safety level required 42 rounds of collision tests. Today, cloud-based virtual collision technologies based on innovations demonstrated by Huawei greatly reduce the number of physical tests required to meet modern crash-safety standards.
Going forward, we expect to have access to the transportation infrastructure technology supplied by the ICT industry for enabling the transformation of automobile routes. In addition to cloud platforms, AI, and Big Data, among the most important of these new technologies are 5G and V2X. This is the future, and these are the new economic growth points of the supply-side revolution that the traditional manufacturing industry does not possess.
FAW’s product development plan has three main directions. The first are semi-autonomous smart, safe cars. The second will be networks of smart cars that work together to alleviate traffic congestion in densely populated cities. The third will be fully automated smart cars that will be certified to operate in smart cities and other specially designated areas.
Additionally, FAW aims to build a cloud platform for Big Data analytics. In the past, automotive enterprises only cared about cars, but now we have to care about roadways, the environment, and an expanding universe of interactions. This digital transformation is an important test for FAW’s R&D capabilities as it will require that our R&D personnel master many new practices and operating regimes that include environmental assessment and AI-based decision-making control, among others.
In contrast, the underlying base of traditional automotive enterprises is mechanical, from the engine and transmission to the integrated electrical components, such as electronic engine control. In the future, the support architecture for the automotive industry will change. A new core platform will emerge that is independent of the traditional engine, transmission, braking, and steering systems. It will include sensors and software intelligence that connect with GPS mapping, sensor fusion, AI, and an increasing variety of supercomputing platforms. This ICT-based architecture will bring a revolutionary change in transportation.
The onboard network architecture for smart cars is equally important. Every car has an electrical system, but traditional wiring is far from meeting smart cars’ requirements. In addition to a sophisticated internal connectivity, smart cars require an outside connection to the cloud. Consequently, a new interconnection architecture needs to be designed that will enable functions such as monitoring the health and attention of the human drivers. For example, if a driver is too tired to continue driving (since most traffic accidents occur because the driver has become distracted briefly), control of the vehicle will be taken automatically and immediately, guaranteeing the safety of people inside and outside the car, as well as the car itself and surrounding physical property.
Another important core technology is dynamic mapping and the many new technologies and services that will be derived from it. Today, all smart car companies are trying to construct new navigation technologies based on maps. One of the bigger engineering challenges is that fixed maps are not able to meet the early list of functional requirements. What is required is the ability to generate real-time maps that are dynamic and responsive to current conditions. In response to the need, FAW has introduced the concept of AllwayEye, whose core function is the ability of each car to capture data related to its immediate environment and upload that information to the cloud. Then all similarly equipped cars in the immediate vicinity will be uploading and downloading situational information with the cloud. If two cars collide, the connected car can upload information about the accident to the cloud for other cars to use their smart planning resources to route themselves around the event location to avoid traffic congestion, as well as having a data record to reconstruct the accident for future reference.
The automotive industry has reached an important crossroads in their digital transformation, one where every manufacturer faces a series of new challenges. How do we choose the best core technologies? How do we transform our traditional R&D models? How do we construct a massive, innovative architecture that is able to respond to an increasingly complex technological revolution?
No single car enterprise can answer these questions based on its own capabilities, and that is why FAW is collaborating closely within the ecosystem of Huawei partners. Our goal is to discuss how to cooperate more and further, as FAW upholds the ideals of collaborating hand-in-hand with partners from all industries to build a car digitalization universe that fulfills our collective vision for the coming smart car era.
FAW Qiming IoV platform
High scalability and flexible choices of T-Box vendors and upper-layer application vendors, and a decoupled architecture provide standard APIs for upper-layer applications. This shields differences of data and commands of different vehicle models, makes it simpler to develop upper-layer applications, simplifies the access of T-Box from different vendors, and accelerates new service rollouts.
FAW-Volkswagen OpenStack development and testing cloud
Open-source Kernel Based Virtual Machine (KVM) at the core, elastically scaling and automatically provisioning IT resources for agile service rollouts. Compatible with other heterogeneous platforms, facilitating the management and O&M of distributed data centers across platforms and geographies and setting up a distributed disaster recovery system across geographies for lowered management and O&M costs.