The automotive industry is undergoing a profound transformation, driven by the convergence of various technological advancements, including software, artificial intelligence (AI) and electric vehicles (EVs). Among these trends, the concept of a Software-Defined Vehicle (SDV) stands out as a game-changer, heralding a new era of connected and intelligent mobility.

At its core, an SDV is a vehicle whose functions and features are primarily enabled through software, rather than relying solely on hardware. This change marks a paradigm shift from the traditional automotive approach, where hardware limitations often dictated the capabilities of a vehicle.

What does “software-defined” mean?

Software-defined is a term that has been around for many years, but it is becoming increasingly important as hardware becomes more standardized (or cheapened) for mass production while software becomes the focus of innovation. True software-defined products are those where the software is the primary driver of functionality and the hardware is simply a commodity, a container, the support.

Since software can easily be updated, a feature that hardware does not share, software-defined products can offer agility, scalability, and redundancy. They are transforming the way data centers are designed and operated. The key to identifying truly software-defined products is to look for those that have abstracted the hardware away and focused on providing solutions through software. These products are the future of IT infrastructure.

To give a simplified example, it is the case of Software-Defined Radios or SDR, where mixers, filters, amplifiers, modulators and demodulators -all these components typically found in the hardware world- are implemented by means of software based on signal processors.

Why the Need for an SDV?

Traditional vehicles are mainly defined by hardware, with software playing a secondary role. This approach has drawbacks, as it limits the ability to introduce new features or adapt to evolving customer demands. As vehicle systems become more complex, the need to centralize data and control of these Electronic Control Units (ECUs) becomes apparent. This is due to the inefficiencies of the current communication system, the redundancy of operations across ECUs and the imbalance of processing power.

While a hardware-defined architecture offers high efficiency, by using Integrated Circuit (IC) hardware-accelerators optimized to execute a specified set of operations
efficiently and with minimal energy consumption, it’s a static solution that cannot evolve to meet changing needs or technological advancements.

In contrast, software-defined architectures provide a more dynamic structure, enabling ongoing updates and improvements throughout the vehicle’s lifespan. This is achieved through Over-the-Air (OTA) updates, which are far more cost-effective and efficient than traditional hardware upgrades that require physical visits to workshops. OTA updates minimize human intervention and reduce the associated expenses, making them an ideal solution for the dynamic and evolving needs of modern vehicles.7

Why all the buzz about the SDV?

So, why are people so excited about software-defined vehicles? By relying on network functions abstracted from the hardware where these are running on, they bring a whole new level of personalization to your driving experience. These vehicles can adapt to your preferences and driving style, using real-time data and machine learning. This means you get personalized infotainment options, optimized navigation routes and even customized comfort settings, promising enhanced safety and continuous innovation.

Talking about safety and security, advanced technologies, like driver-assistance systems and autonomous driving capabilities, are used to significantly reduce accidents and make our roads safer. Equipped with sensors, cameras and smart predictive software, these smart vehicles can detect and respond to potential hazards, preventing collisions and ensuring road safety. On top of that, they’ve got strong cybersecurity measures in place to protect your vehicle data and keep you safe from cyber threats.

This innovative approach allows for simultaneous development in both the physical and digital worlds, empowering software to drive differentiation. And beyond mere technological siren calls, they stand as catalysts for sustainability and economic growth, optimizing the life cycle and value-cycle of vehicles. This double effect not only makes driving better but also shows how this new conceptualization of the car play a part in making our future more eco-friendly and economically strong by being part of an interconnected ecosystem, exchanging real-time data and communicating with smart city system to optimize fuel efficiency, reduce emissions and seamlessly embrace electric vehicle technology. Plus, they’re not just vehicles; they’re job creators bringing more opportunities in software development, data analytics and connected vehicle services.

Related content: Innovations in Control Systems: A Deep Dive with Industry Engineers 

Potholes on the road of Software-Defined Vehicles

Now we know that Software-Defined Vehicles represent a groundbreaking shift in the automotive industry. However, the road to this future is not without its challenges.

“With great power there must also come great responsibility”. And this is the case with moving from dedicated hardware to generic platforms with customizable software. Which comes at a cost of complexity and responsibility of software development. Unlike traditional vehicles with relatively simple software architectures, SDVs are equipped with complex software systems that control various vehicle functions, from navigation to autonomous driving. This complexity increases the risk of software bugs, latencies and cybersecurity vulnerabilities, demanding rigorous testing and validation processes.

The increasing connectivity and data-driven nature of SDVs make them potential targets for cyberattacks. Robust cybersecurity measures are essential to protect sensitive vehicle data, prevent unauthorized access and safeguard against cyberattacks that could compromise safety and privacy.

In this context, these attacks could range from stealing your home address to scenarios like introducing non-existent objects on the road to cause the vehicle to malfunction or manipulating the perceived traffic to guide the vehicle along a specific path for targeted attacks. And with the expansion of connectivity ensuring compatibility between different software components and hardware systems becomes a critical issue.

This challenge is accentuated by the fragmented nature of the automotive industry, with many automakers and suppliers developing their own unique systems and standards. Addressing compatibility issues requires automakers to adopt open standards and collaborate with third parties to develop a unified framework for software development and interoperability, as well as common integration and testing techniques to guarantee a seamless integration and reduce the risk of software transient glitches.

The transition also presents challenges for the automotive workforce. The demand for skilled software engineers, data scientists and cybersecurity professionals is growing rapidly and automakers need to invest in up-skilling and retraining their workforce to meet these demands. Additionally, new business models and partnerships will emerge as the new automotive industry evolves, requiring a new mindset and approach from automakers.

The 2020’s Chip Shortage: A Blessing in Disguise?

The persistent semiconductor shortage has significantly disrupted the automotive industry,leading to production delays and price increases. Interestingly, this shortage could inadvertently boost the transition towards SDVs. As car manufacturers are struggling to get enough chips for conventional vehicle designs, they might be more motivated to investigate newer technologies that require fewer specific chips.

A recent analysis on Tom’s Hardware from January 23, 2024, indicates that TSMC appears to be prioritizing the production of smaller quantities of newer, more costly chip technologies over mass-producing older technology-based chips, which offer a lower profit margin. This shift in production strategy could have implications for industries that rely on these chips, including the automotive sector, and could help with the scenario mentioned earlier.

Architecture approaches: From vehicle to cloud

Different automotive companies are adopting varying approaches to SDV architecture. Some are developing proprietary architectures, while others are collaborating with software giants like Google and Microsoft to create standardized platforms. Each approach has its own advantages and disadvantages, and the industry is still in the early stages of determining the most effective architecture for SDVs.

Electric perspective

Automakers are actively exploring diverse architectural strategies to advance the maturity of software-defined vehicles, but the tendency is to aim for a Central or Zonal Architecture. In this framework, the vehicle is strategically divided into multiple zones, as opposed to conventional architectures with dedicated hardware components for specific functions. In contrast to the traditional approach of bolting on dedicated hardware for every function the Zonal Architecture seeks to ditch the maze of wires and specialized computers and, instead, dividing the car into smart zones, each controlled by a powerful central hub. The main objective is to simplify the vehicle’s design by reducing hardware complexity and making more efficient use of software.

This new conceptual architecture for the automotive world addresses this issue by grouping functions into zones based on their location or function. For instance, a singular ECU or “Zonal Gateway” might oversee all functions related to the vehicle’s front end, while another is responsible for managing all in-cabin functionalities. By centralizing control within designated zones, automakers enhance the overall efficiency and adaptability of the vehicle’s electronic
systems, contributing to a more efficient and easily maintainable design.

Data-driven perspective:

The evolution of vehicle architectures in the next generation paves the way for consolidating specialized control units into versatile edge systems, showcasing elevated performance capabilities. These systems not only facilitate extensive data processing and communications but also signify a departure from traditional monolithic software structures toward contemporary architectures. The transition taps into not just the volume of data but also the content within it. Introducing the cloud domain into the equation is a key aspect of this transformation. This means that data processing is no longer confined solely to the vehicle domain; it extends beyond, allowing real-time consumption without delays.

In this cutting-edge scenario, the cloud domain complements the in-vehicle development infrastructure. The typical in-vehicle setup includes a robust centralized computer capable of handling a flexible mixed-critical workload, adhering to Automotive Safety Integrity Levels (ASIL). Accompanied by a service-oriented middleware, this infrastructure manages local Real-Time Operative System (RTOS) processes coexisting seamlessly with microservices in the cloud.

Microservices, as individual processes customized to specific capabilities, offer independent scalability. This transition from a monolithic to microservices architecture significantly enhances the reusability and robust updating of complex vehicle software components.

From edge computing to cloud infrastructure, numerous companies are strategically aligning themselves, each presenting a unique blend of ingredients. It comes as no surprise that industry leaders, like Nvidia and Qualcomm, well known for their processors, are developing solutions that bring computing closer to the vehicle, enabling real-time data analysis and rapid decision-making. On the other hand, companies such as Bosch, Continental and Red Hat Linux, are prioritizing the development of microservices and cloud infrastructure, without overlooking the essential role of silicon in the overall architecture.

While various architecture proposals are under consideration, a consensus is gradually forming, aligning with the emerging structure in the automotive landscape. From my perspective, a generic SDV software architecture with cloud infrastructure in mind can be outlined through five distinct layers, as detailed in the table below:

Layer Description
AI Products Tools and platforms for software development, AI engineering, data management, network automation, and connected vehicle use cases.
Security Protects SDVs from external and internal threats, adopting a zero-trust framework for data and communications.
AI and Data

Platform

Enhances SDV development with AI models and data analytics. AI is used for various functionalities, including ADAS, cybersecurity and telematics data analysis. Generates artifacts like test cases and software source code using Generative AI.
Hybrid Cloud

Platform

Uniform Linux and Kubernetes-based platform, enabling flexible distribution of software containers from the vehicle to backend systems.
Infrastructure Includes vehicle, telco equipment, roadside units, smart city systems, and OEM RTOS and backend systems, facilitating continuous data exchange.

The Future of Software-Defined-Vehicles

The transition to SDVs is still in its early stages, but the potential benefits are significant. As the technology matures and costs decrease, we can expect to see a rapid acceleration in the adoption of SDVs. This will transform the automotive industry, creating new opportunities for software developers, automakers and transportation providers.
Despite the challenges, the Software-Defined Vehicle concept holds immense promise for the future of mobility. By adopting software-centric developments, automakers can create vehicles that are more intelligent, connected and adaptable to changing user needs. In a near future, we can expect to see a wave of SDVs hitting the roads, transforming the way we drive and interact with our vehicles. With the entrance of new players, such as Xiaomi and Sony Honda Mobility (SHM), the industry is undergoing a notable diversification, further reshaping the dynamics of innovation and competition.

The future of mobility is software-defined, and the journey is just beginning.