Automobile manufacturing is the pillar industry of industrial manufacturing in the past hundred years, with the arrival of the era of automobile electrification and intelligence, the automotive industry is facing a huge change, and a new consensus has been formed: when the automobile gradually evolves from a hardware-driven mechanical and electrical device to a software-driven electronic product, the competition law of the automotive industry will be rewritten.
In the future, software is expected to surpass traditional hardware to become the core of the value of the whole vehicle.
Industry leaders are adjusting their course, moving from hardware-centric to software-centric in terms of investment direction, and on this basis, launching new on-demand services and opening up new revenue streams. This will lay the foundation for a new generation of software-driven solutions that will drive change and long-term growth in the automotive industry.
To this end, automakers are embarking on a restructuring journey stemming from digital transformation, which has prompted automakers to introduce new information and operational technologies to help develop the strategy of the "new four modernizations" (vehicle connectivity, automation, sharing and electrification) of the next generation of vehicles. Software is the foundation of digital transformation, and software-defined cars will be the gateway to the mobility of the future. Automakers will unlock new business models through software-defined approaches, create new revenue streams, and deliver personalized customer experiences. As the driving experience and customer expectations change, what was once high-end features will become standard. If existing automakers can't meet changing customer needs, new carmakers who offer convenient, economical, and efficient alternatives will take more market share.
How is software disrupting the automotive industry?
Until now, the automotive industry has been able to foresee the scale of software change in the industry. Hardware used to be the main focus of automotive innovation, but it is software that now attracts attention and investment. Frost & Sullivan predicts that by 2030, automakers (OEMs) will spend around $2.3 billion on connected vehicles and autonomous driving software infrastructure.
But if the complex electrical and electronic architecture of the current generation of vehicles remains the same, is it possible for a software-defined car to be achieved? How do we cultivate the human resources needed to develop related technologies? How will cloud computing help accelerate the evolution of software-defined vehicles? Without the right expertise, addressing these issues can take a significant amount of time and resources into the development cycle.
In addition to unlocking new safety, comfort, and convenience features, software-defined cars have several advantages over their hardware-defined predecessors. Currently, remote software updates (over-the air updates -- OTA) provide limited software upgrade services that can only help solve a small number of software problems, and vehicles still need to contact the dealer for on-site troubleshooting after a diagnostic fault.
"Software-defined cars, automakers will seamlessly develop and deploy new software services, including information acquisition, improved safety, and changes to the vehicle's powertrain, driving power, range of electric vehicles and other functions. This requires automakers to move away from the siloed mentality of treating vehicles and cloud computing separately and envision an edge-to-cloud approach. ”
This edge-to-cloud approach will be a key premise and driver for the development of software-defined vehicles. In a software-defined future, automakers need trusted technology partners who have both the ability to explore innovative business models and the tools and talent to implement them.
The automotive industry also faces significant challenges as it strives to transform, including:
· A large number of legacy technologies – for example, today's electrical and electronic architectures and road computers operate independently and cannot be scaled; lack the ability to acquire data related to hardware consolidation and provide normalized and secure data access. As a result, developers cannot create new micro-services.
· A container approach that deploys new features for cross-domain update services and development is needed to develop containers in the cloud and deploy them in-vehicle.
· The flexibility of the software experience, the challenge for car companies is how to integrate modern continuous integration and continuous deployment (CI/CD) tools and frame driving with the existing "V" development model.
How to maintain innovation when the threshold for car building is rising rapidly?
Software-defined car solutions and outcomes based on Amazon Cloud Technology
Seamless development and deployment
· Build a simplified architecture that decouples hardware from software and enables seamless interaction between vehicles and the cloud, dramatically shortening the full cycle of architectural development and software development around vehicles. This means that automakers will spend significantly less resources on vehicle architecture and accelerate time to market.
· With the combination of continuous integration and continuous deployment (CI/CD) and automotive DevOps culture, challenges such as the integration and implementation of remote software updates (OTAs) can be addressed by deploying containers (including software upgrades, troubleshooting, and new applications) in a more agile environment. The entire development process of software being written, tested and deployed in a vehicle will undergo a major change.
Machine learning at the edge
Pushing data acquisition, storage, machine learning models, and deployments that were done in the data center back into the vehicle, creating data insights and machine learning experiences in the vehicle requires as much hardware and software decoupling as possible across vendors and Tier 1 parts vendors, so the two results cannot be isolated from each other. Regardless of vehicle make/model, only a cross-vendor format will get the most out of the generalized vehicle development platform in the coming years.
Amazon Cloud Technologies and its partners are driving software-defined cars
· Amazon's cloud technology distributed architecture from the edge to the cloud can be seen as a three-pronged approach to the whole. The first is the vehicle itself, which requires a generalized platform that provides in-vehicle computing and integrates with managed IoT services.
· This is followed by the network edge, where Amazon Cloud Technology expands compute storage at the edge of the network through integration with telecom service providers. This enables ultra-low latency compute, storage, and startup of IoT services.
· Finally, the cloud platform is flattened, which offers a lot of flexible, scalable and on-demand services such as the Internet of Things, machine learning, etc. In addition, edge services help customers run cloud services/instances outside of Amazon's cloud infrastructure.
Amazon Cloud Technology offers a variety of services, but it follows an ecosystem "first" approach, working with a wide range of partners and ecosystem players to deploy solutions from the edge to the cloud. One example is Continental Automotive Edge (CAEdge), which deployed a virtual workbench on Amazon Cloud Technology, providing a range of tools for intensive software development and maintenance to meet the needs of OEMs and related partners for software-intensive vehicle architecture development environments, and to improve development/test efficiency and securely deploy OTA directly to vehicles.
Amazon Cloud Technologies, on the other hand, is also partnering with industry-wide through cloud-native-based SOAFEE (Embedded Edge Scalable Open Architecture). Amazon Cloud Technologies continues to advance its own and partners' solutions in software-defined vehicles in the following four key areas.
Democratizing data and simplifying the developer experience: Amazon Cloud Technologies and BlackBerry QNX have jointly developed a smart vehicle data platform, Blackberry IVY. The platform enables automakers to share insights with a large community of developers, where developers don't need specialized automotive skills to create machine learning-powered features and situational-aware experiences. In addition, customers can leverage other parts of Amazon's cloud technology portfolio to provide deeper business and operational analytics, as well as the ability to deepen and enhance data lakes.
Automated DevOps: From the edge to the cloud, automated DevOps plays an important role. Flexible management of continuous integration/continuous deployment (CI/CD), along with a mix of critical container orchestration, environment parity, and application-layer networking, supports an agile, modern, automated DevOps environment. As the electrical and electronic architecture is updated in the next generation of vehicles, this approach will provide a direct foundation for software-defined vehicles.
Hardware architecture and design: Comprehensive electrical and electronic architecture with high-performance computing (HPC) and enhanced networking capabilities will support hardware and software decoupling. This will help reduce wiring costs and reduce vehicle weight. With a holistic architecture from edge to cloud, ARM simplifies the development process by performing environment parity through its CPU deployed in the car (e.g., NXP S32g) to ARM instances deployed on Amazon Cloud Technology. By using Amazon Cloud Technology Graviton, performance was boosted by 30% while significant cost optimization was achieved.
Hardware and software abstraction layer: Hardware and software abstraction plays an important role in the software-defined car, and the Apache TVM and Greengrass V2.0 open source services provided by Amazon Cloud Technology are decoupled with the service support provided by Tier 1 vendors. The idea behind software abstraction and orchestration techniques is to run containers and AL/ML instances at the edge using Greengrass and Amazon SageMaker Neo's integrated approach to make it easier for customers to implement in a more compact way.
Amazon Cloud Technology brings all of this together by providing its Tier 1 partners with a way to build their solutions, and for that, they can use Greengrass, a general-purpose and foundational computing service. Manufacturing, IoT services, and automotive customers can seamlessly connect their edge devices with any Amazon cloud technology/third-party service.
With its unique innovation DNA, Amazon Cloud Technology has become an integral part of customers' strategies to reduce time to market. As shown by NXP, an ARM in-vehicle CPU that has been integrated with Greengrass, Amazon Cloud Technology takes an ecosystem-based approach, and the combination of NXP S32G and Amazon Cloud Technology allows customers to deploy cloud-native applications and gain access to machine learning use cases.
What does software-defined mean for the automotive industry?
The transformation of electrical and electronic architectures and the associated development of vehicle manufacturing require closer partnerships between automakers, suppliers and technology suppliers. Amazon Cloud Technologies, together with its partners, is creating the comprehensive platform the industry needs that extends vehicles to the cloud while reducing the complexity of software development and system integration. This opens up the prospect for automakers to improve lifecycle management and develop innovative capabilities that can be delivered to customers, all of which will build deeper, closer relationships with customers.
Software-defined cars will create a variety of opportunities for consumers and OEMs, many of which are not even yet conceived. Frost & Sullivan believes that by 2025-2026, some form of software-defined use case will emerge for vehicle projects and will continue to expand, advancing new business models and data monetization opportunities. Amazon Cloud Technologies is designed to provide secure, scalable, and cost-optimized services for future software-defined automotive applications. Amazon Cloud Technology is committed to creating a single platform that spans different fields, different brands and different models. This will enable OEMs to invest in strategic platforms/use cases, allowing them to focus on pushing the boundaries of experiential customer service.
For more excitement, please pay attention to the Amazon Cloud Technology "Software-Defined Car" online conference on April 21, 2022, 14:00-16:00.
Scan the QR code below to register for an appointment: