Classic AUTOSAR vs Adaptive AUTOSAR

Can Classic and Adaptive AUTOSAR coexist and complement each other?

Classic AUTOSAR is not obsolete; it remains essential for real-time ECU software, but it needs a complementary platform – Adaptive AUTOSAR – to meet modern requirements for connectivity and frequent updates. In other words, the two platforms are designed to coexist, each addressing different but increasingly interdependent needs in future vehicles.

In 2002, AUTOSAR was introduced to standardize the development of automotive software for ECUs. At that time, ECU software was assumed to be deeply embedded and not expected to require updates throughout the vehicle's lifetime. Classic AUTOSAR was therefore optimized for stable, highly predictable real-time behavior in tightly controlled environments.

Since then, the automotive software industry has changed dramatically, with vehicles now requiring frequent software updates and continuous connectivity. The Classic AUTOSAR architecture, built on a robust real-time embedded operating system, can experience delays and interference when the car is connected to external networks, which can hinder strict real-time communication.

This limitation does not mean that Classic AUTOSAR should be replaced. Instead, it highlights the need for a companion platform that can handle dynamic, update-intensive, and networked use cases. Adaptive AUTOSAR was created precisely for this purpose, allowing both platforms to coexist and jointly address the evolving demands of future vehicles.

What is Classic AUTOSAR?

Classic AUTOSAR is a standardized set of specifications for building embedded software platforms that implement fundamental automotive functions while abstracting away the underlying hardware. It defines how software components interact with basic software and hardware so that applications can be developed and reused consistently across different control units.

In a multi-ECU environment, Classic AUTOSAR platforms must provide BSW (Basic Software) and RTE (Runtime Environment) abstractions that allow OEMs and suppliers to implement essential applications without dealing directly with hardware details. The Basic Software layer offers key services to application developers, including operating system services, network communication and management, memory handling, and diagnostics.

Because of these abstractions, individual software units are largely independent of the specific hardware environment, and developers typically do not need to be aware of it, as illustrated in Figure 1.

What is Adaptive AUTOSAR?

Adaptive AUTOSAR is a flexible, service-oriented software platform designed to support both basic and advanced cross-domain functions on modern multi-core, multi-ECU and high-performance computing (HPC) architectures. It extends beyond traditional automotive ECUs to enable connected, updateable and dynamically deployable vehicle functions.

An Adaptive Platform differs from a Classic Platform because it can handle both basic and advanced cross-domain functions on top of multi-core, multi-ECU and HPC-integrated environments. Advanced functions include support for communication with both on-board and off-board systems, use of cross-domain computing hardware, over-the-air (OTA) updates, remote repair and exchange handling, and the dynamic deployment of customer functions, including driving functions.

Due to the segregation of software and hardware in Adaptive AUTOSAR, tasks are distributed differently between OEMs and suppliers. Instead of ordering a functional block as a physical device in the vehicle, it becomes possible to purchase only the software. By installing apps from an app store, the driver can effectively take on the role of a software integrator.

An illustration of the AUTOSAR Adaptive Platform Architecture typically shows the different layers of the platform and how they interact.

What are the differences between Classic AUTOSAR and Adaptive AUTOSAR?

Classic AUTOSAR and Adaptive AUTOSAR differ mainly in their operating system base, communication style, programming language, scheduling, real-time performance, safety targets, and flexibility. Classic AUTOSAR is optimized for resource-constrained, hard real-time ECUs, while Adaptive AUTOSAR targets high-performance, flexible platforms that support over-the-air updates, IoT integration, and autonomous driving functions.

The key differences between Classic AUTOSAR and Adaptive AUTOSAR are summarized in the table below:

Table 1: Comparison between Classic AUTOSAR and Adaptive AUTOSAR.

While the classic standard has supported the fundamental automotive functions of the past and present, upcoming functions related to the Internet of Things (IoT) and autonomous driving will primarily rely on the adaptive standard.

How Can Classic and Adaptive AUTOSAR Coexist in One Vehicle Architecture?

Classic AUTOSAR and Adaptive AUTOSAR are designed to coexist and collaborate within the same vehicle architecture, not to replace one another. Each platform is used for different types of applications, and communication between them is enabled via a gateway ECU that bridges traditional vehicle buses and Ethernet, including the required signal and service conversions.

The Classic AUTOSAR platform is well suited to applications that demand high safety and strict real-time behavior, so embedded software functions with these requirements are typically deployed on this platform. In contrast, the Adaptive AUTOSAR platform is designed for parallel processing of large data volumes and high-performance computing, making it appropriate for such functions.

As autonomous driving, IoT, and cloud technologies continue to evolve, Adaptive AUTOSAR is emerging to address both current and future automotive technology needs. With its flexible architectures and complementary features, it can support a wide range of adaptive deployments, more complex microcontrollers, and interactions with various non-AUTOSAR systems.

Adaptive AUTOSAR also introduces a substantial change to vehicle architecture by relying on Ethernet as its communication network. While Adaptive ECUs use Ethernet, Classic ECUs continue to use traditional in-vehicle bus networks such as CAN or LIN.

To enable communication between Classic and Adaptive ECUs, a gateway ECU is used. In this setup, a Classic ECU can act as a gateway by packaging signals from the bus system into a service format that an Adaptive ECU can interpret.

A configuration format conversion is required in this process. If the Classic ECU is purely signal-based and cannot package signals directly into services, it converts the signals into UDP frames and transmits them over Ethernet. The Adaptive ECU then uses a "signal to service" mapping capability to transform these UDP frames into services it can consume.

What Does The Future Hold For AUTOSAR?

The Classic platform is designed for functional ECUs that run deeply embedded software, while the Adaptive platform is focused on enabling autonomous driving and other advanced functions, representing a more future-oriented approach. Both platforms remain equally essential to the automotive industry.

It is possible that the Adaptive AUTOSAR platform may ultimately become the primary or even the only platform used for ECU software development. However, it is still too early to make a definitive prediction.

For now, automotive engineers can bring these two worlds together by using flexible gateways between Classic and Adaptive systems, allowing them to combine their strengths and maximize overall potential.

Conclusion

Classic and Adaptive AUTOSAR are not rivals but complementary answers to a rapidly evolving automotive landscape, with Classic securing deeply embedded, real-time and safety-critical functions while Adaptive unlocks connected, high-performance, and updateable capabilities. By leveraging different operating systems, communication styles, and deployment models, they together enable everything from deterministic microsecond control to service-oriented, over-the-air enabled features that span vehicle and cloud. Their coexistence is made practical through gateways that bridge CAN/LIN and Ethernet, translating signals into services so each ECU type can play to its strengths. Looking ahead, as autonomous driving and IoT mature, the real advantage will belong to teams that learn not only to choose between Classic and Adaptive AUTOSAR, but to artfully orchestrate them as one coherent software-defined vehicle platform.

Reference

The following official AUTOSAR resources provide more detailed information on the Adaptive Platform and Classic Platform:

• AUTOSAR Adaptive Platform standard: https://www.autosar.org/standards/adaptive-platform
• AUTOSAR Classic Platform standard: https://www.autosar.org/standards/classic-platform