A Benchmarking Guide for Software-Defined Vehicles

S&P Global Mobility has developed a method which we refer to as SDV Readiness to help the industry and consumers alike assess SDV technology.  

We started with what constitutes an SDV — namely, a vehicle that is fully updatable and upgradeable over the air across all domains throughout its lifetime, made possible by the decoupling of hardware and software.

Importantly, SDV Readiness refers to the vehicle’s capabilities—not the readiness of the automaker or its suppliers — allowing for an objective evaluation of its technological foundation and ability to support SDV features over time.

After gathering feedback from OEMs, software suppliers, chip makers, electronics suppliers and cybersecurity experts, we developed this method to objectively categorize and benchmark SDV readiness levels. 

To build the benchmark, we focused on three foundational technologies, or building blocks, enabling SDVs:

• Over-the-air (OTA) capabilities;
• Hardware (specifically the E/E architecture); and
• Software components.

Our analysis is grounded in 35 technical parameters tracked at the vehicle level across these technology categories, all of which help enable smooth, cost-efficient, multi-domain, frequent and secure vehicle upgrades. These parameters cover domains covered by OTA, Service Oriented Architecture, Unified Operating Systems, deployment of zonal EE Architecture, communication protocols, and Cybersecurity.  

Optimizing these factors helps manufacturers reduce initial development and update costs, minimize downtime and enhance the overall user experience.

The SDV Readiness Level benchmark helps automakers answer pressing questions such as:

• Which OEMs are most ready for SDVs?
• Which ones are lagging?
• Can traditional OEMs catch up with EV startups and Chinese brands?

OEMs can see how they rank in comparison to competitors, from Level 0 (Not Connected) to Level 5 (Fully Ready). Each level represents a different stage of technological advancement and carries unique implications for OEMs and their end customers, while also being defined by specific technologies that enable their capabilities.

Technologies Defining Level 0:

• Basic E/E Architecture: The reliance on a traditional CAN bus system limits communication speed and flexibility.
• Static Software Components: Software is tightly coupled with hardware, making it difficult to implement updates.

At Level 0, vehicles operate on a basic signal-based architecture. For OEMs, this means limited scalability and increased costs, as the CAN bus backbone restricts data transfer speed and flexibility. Software remains static and fragmented, necessitating manual interventions for updates or bug fixes. Cybersecurity measures are minimal, leaving vehicles vulnerable to threats.

For OEMs, the challenge lies in navigating the operational costs associated with physical service interventions for updates, which ultimately impacts customer satisfaction and brand reputation.

Technologies Defining Level 1:

• OTA Capabilities: Vehicles can receive updates, but only for existing functionalities.
• Limited Cybersecurity Measures: Protection is primarily at the ECU level, focusing on infotainment and connectivity.

Level 1 vehicles gain the ability to receive over-the-air (OTA) updates, albeit limited to refreshing existing data, such as navigation maps. For OEMs, this marks a step forward, yet the potential for revenue generation remains restricted to subscriptions for map updates. Cybersecurity measures at this level are still minimal, primarily focusing on infotainment and connectivity.

While customers benefit from having the latest map data, the inability to install new applications necessitates manual maintenance interventions. This level enhances the user experience but underscores the need for OEMs to develop more robust capabilities to fully leverage connectivity.

Technologies Defining Level 2:

• Service-Oriented Architecture (SOA): Some vehicle domains are on SOA, which enables modular development, allowing for dynamic updates and improved integration. However, majority of domains remain on Signal-based software architecture.
• Enhanced OTA Capabilities: Vehicles can now receive updates for both software and application code, allowing for bug fixes and new functionalities.

At Level 2, OTA capabilities extend beyond data refreshes to include upgrades for autonomy or powertrain applications. This gradual transition to a SOA allows for greater software modularity and scalability, reducing development costs and time-to-market. This shift enables manufacturers to respond more rapidly to market demands.

This level allows OEMs to enhance the vehicle experience, as a smarter and more connected vehicle enables real-time navigation, predictive maintenance, and AI-enhanced driver assistance. Continuous software upgrades also ensure compliance with evolving regulations.

Technology Defining Level 3:

• Zonal E/E Architecture: This architecture reduces complexity by streamlining communication between electronic control units (ECUs) and centralizing control.

Level 3 introduces zonal architecture. Migrating from distributed or domain to zonal E/E architecture reduces vehicle complexity, manufacturing costs, and weight. Zone controllers also enable to update or upgrade multiple functions and domains with the same OTA, cutting development time and cost for OEMs. 

Technologies Defining Level 4:

• Cloud-based Cybersecurity Framework: A unified, cloud-based operating system (OS) enhances security measures across all vehicle domains.
• Cross-Domain OS: Cross-domain OS’s enable real-time data sharing between ADAS, infotainment, and powertrain domains.
• Improved OTA Capabilities: OTA is available for three or more domains. More frequent updates ensure that vehicles remain equipped with the latest features and security patches.

Transitioning to Level 4, OEMs adopt Ethernet-based networks to manage the increasing data demands of modern vehicles. This shift streamlines software development and enhances cybersecurity through a cloud-based centralized framework.

By implementing a unified OS, manufacturers can also reduce complexity and improve integration. Migrating from individual OS to a cross-domain OS streamlines software development, reduces hardware complexity, and improves integration across vehicle functions. A unified cross-domain OS is less expensive to upgrade and OEMs can create a scalable software platform that accelerates development across multiple vehicle models.

Additionally, computing power and memory can be dynamically allocated across different domain functions, optimizing system performance and avoiding underutilized hardware, leading to better efficiency and lower overall power consumption.

Technologies Defining Level 5:

• Advanced Zone Controllers: These controllers manage multiple functions efficiently, enhancing vehicle performance and reducing complexity.
• Unified Operating System: This architecture supports cross-domain function integration, enabling predictive maintenance and AI-driven optimizations.
• Ethernet-Based Networks: These networks allow for higher data transfer speeds and better management of complex vehicle functions.

At Level 5, vehicles reach a pinnacle of capability, using zone controllers that consolidate functions into a single point, improving manufacturability and reducing wiring costs. A unified OS reduces software fragmentation. This level allows OEMs to standardize software across different models, accelerating development cycles and reducing operational costs.

As vehicles evolve with technological advancements, the architecture supports cross-domain function integration, allowing for predictive maintenance and AI-driven optimizations. This capability reinforces OEMs' positions as leaders in the automotive market.

From the consumer’s perspective, reaching Level 5 enables the creation of a seamless digital ecosystem where the vehicle becomes a natural extension of the user's connected life. The car can dynamically link with the cloud, backend services, and the user's personal devices — like their smartphone, smartwatch, or home assistants.

This connection allows the vehicle to learn and reflect user preferences, driving styles, and lifestyle patterns. It means a deeply personalized experience: from preferred seat and climate settings to curated infotainment, navigation suggestions, and even proposing the most relevant applications, services, or subscription packages tailored to the user's habits.

On the OEM and industrial side, a Level 5 SDV-ready vehicle opens the door to continuous data collection — always within legal frameworks like GDPR — that fuels machine learning algorithms and predictive analytics. This data not only improves the in-vehicle experience but also feeds back into the entire factory and supply chain ecosystem.

• Development teams can better understand real-world usage patterns and adjust designs faster.
• Factories can optimize production lines based on real operational data, making continuous improvements without waiting for major model refreshes.
• Supply chains can become more responsive and predictive, minimizing disruptions and aligning parts production and delivery with actual usage and needs.