Software in Automotive: Top OS Requirements for the Software-Defined Future
The Internet of Things (IoT) trend is moving fast across sectors. Software-defined systems in healthcare, industrial, smart cities, and transportation are driving innovative electronic architectures, powerful chips, and operating systems. The software-defined vehicle (SDV) is generating buzz among auto OEMs (original equipment manufacturers) and their technology partners for the new capabilities it enables, including advanced safety features, autonomous vehicles, vehicle-to-everything connectivity, over-the-air upgrades, on-demand functions, and data-driven services.
The foundation of the SDV is the hardware in-vehicle electronic architecture. Over the next 15 years, in-vehicle architectures will migrate from distributed functions and domain controllers to zonal controllers. This evolution enables higher complexity in electronic systems, vehicle weight and wire cost reduction, and greater flexibility and safety. The zonal controllers also take advantage of advances in high-performance computers based on powerful SoC (system-on-a-chip) devices.
Top OS Requirements for In-Vehicle Architectures
Playing a critical role in this new world are embedded operating systems that are also evolving to support more powerful in-vehicle computing and more demanding in-vehicle use cases. In flat distributed architectures, the electronic control units (ECUs) need an OS that manages a variety of programs that control the hardware and the applications each ECU is designed to accomplish.
As the in-vehicle architecture evolves and becomes more centralized, the multiplicity of the software increases; this in turn increases the complexity, safety, and security needs of the underlying OS. Having an OS that addresses the next-generation architecture is critical to maximize all the benefits of it — including maximizing performance of the system while maintaining stringent safety and security measures. The requirements for the next-generation operation systems that fit zonal in-vehicle architectures are:
- Real-time processing. The ability to complete critical requests fast and in a timely manner. In the event of hardware issues or interruptions, the OS handles it instantly.
- High-performance execution. The ability to execute various requests in parallel due to higher functionality expectations.
- Scalability. The ability to adapt to business change, allowing easier integration of new features and functions throughout the lifecycle without requiring a complete overhaul of the entire system. The OS can execute the exact same way in parallel on 32 cores as it would on 8 cores.
- Low latency. The ability to manage critical tasks quickly and reliably. Applications in medical, transportation, and smart cities are time-critical; one microsecond off could result in a situation where life could be at risk.
- Security. In-vehicle software must be hardened with security capabilities, meet industry certifications, and stay compliant with regulations.
- OS migration. Car manufacturers are starting to consolidate safety-critical and non-safety-critical functions that will be managed by only a single operating system.
- Hardware-agnostic. The OS supports different usages and doesn’t require any hardware adaptation. It can accommodate artificial intelligence and machine learning algorithms as full autonomous driving and GPU (graphic processing units) developments.
The Growing SDV Market Means New Business Opportunities
According to the S&P Global Mobility report, “The Changing Landscape of E/E Architecture,” the first software-defined vehicles with zonal controllers and next-generation software are expected to hit the road in 2025. Battery electric vehicles will be the first to adopt the technology, with hybrids and internal combustion engine vehicles soon to follow.
Software is becoming essential to the design of IoT devices and for SDVs. The automotive ecosystem is migrating from traditional Tier 1 hardware suppliers to software and high-tech vendors. Today, nearly 40% of the total cost of developing a vehicle is related to software and electronic components. Coding lines are increasing with software functionality, and the amount of data produced in the car is growing. Vehicles are, quite literally, becoming mobile data centers that open additional revenue opportunities.
With the advent of the software-defined vehicle, we should expect higher integration across auto OEMs and IoT actors in sectors such as connectivity, edge computing, cloud, and big-data analytics to deliver improved customer experiences, safety, and ROI maximization.