FAQ

General

What is OSxCAR?

OSxCAR (Optimized Software-defined Car Architectures) is an EFRE-funded research project developing an innovative platform for software-defined vehicles (SDV). The project combines WebAssembly technology, AI-powered optimization, and a flexible testbench environment.

Who finances OSxCAR?

The project is funded by the EFRE/JTF NRW – NeueWege.IN.NRW program. Project coordinator: PTJ. Funding code EFRE‑20800271. The total funding amounts to 5 million euros.

How long does the project run?

The project duration spans from June 2024 to May 2027 (3 years).

Who are the project partners?

The consortium consists of five partners:

• Aptiv Services Deutschland GmbH (Consortium leader) – Hardware integration and system architecture
• Bergische Universität Wuppertal – Research and development
• Universität Bielefeld – AI and machine learning
• paraXent GmbH – Low-latency switches and FPGA technology
• CETEQ GmbH – WebAssembly and cloud-based test framework

Technology & Platform

Which technologies are in focus?

• SDVA-Testbench: Runtime-configurable, heterogeneous hardware platform
• WebAssembly (WASI/WIT, AoT): Platform-independent, secure software execution
• AI approaches (GNNs): Optimization of network and software distribution
• RECS Microserver: Heterogeneous computing architecture (x86, ARM, RISC-V)

What is a Software-Defined Vehicle (SDV) platform?

An SDV platform enables flexible, software-controlled configuration of vehicle functions. Instead of hard-wired ECUs, functions can be dynamically distributed and updated as software modules. OSxCAR develops a Level 0-5 framework for SDV systems.

Why WebAssembly in vehicles?

WebAssembly offers several advantages:

• Platform independence: One binary runs on different hardware architectures (x86, ARM, RISC-V, MCUs)
• Security: Strict sandbox isolation protects critical systems
• Performance: Ahead-of-Time (AoT) compilation for deterministic latencies
• Portability: Development on laptop, testing in cloud, deployment on target hardware

What are Behavioral Envelopes?

Behavioral Envelopes define safe operating boundaries for third-party apps in vehicles. They enable dynamic software homologation by guaranteeing that apps only operate within approved parameters (e.g., latency, resource consumption, communication patterns).

Which hardware platforms are supported?

OSxCAR supports heterogeneous hardware:

• CPUs: x86, ARM, RISC-V
• Accelerators: GPUs, FPGAs, ASICs
• MCUs: Microcontroller-compatible WebAssembly containers
• TEEs: Trusted Execution Environments for safety-critical applications

Testbench & Development

What is the SDVA Testbench?

The Software-Defined Vehicle Architecture Testbench is a runtime-configurable hardware platform that simulates various vehicle topologies. It enables:

• Remote access for distributed teams
• Flexible network configuration (Ethernet, CAN, LIN)
• Fast prototype testing without physical hardware duplicates
• Cloud-based sharing with TISAX compliance

How quickly can a test environment be set up?

With OSxCAR, setup time reduces from months to minutes. The runtime-configurable switching matrix enables immediate topology changes without hardware rewiring.

Can I use the testbench remotely?

Yes, the SDVA Testbench is designed for remote access. Teams worldwide can access the test environment, conduct experiments, and collect data – with time-slot reservation and secure cloud connectivity.

AI & Optimization

How is AI used in OSxCAR?

AI models (especially Graph Neural Networks, GNNs) are used for:

• Latency prediction: Forecasting network delays
• Software placement: Optimal distribution of functions on compute nodes
• Resource allocation: Intelligent load balancing
• Routing optimization: Efficient network path selection

What accuracy do the AI models achieve?

Initial GNN models (e.g., RouteNet) achieve single-digit percentage accuracy for latency predictions. The models are continuously improved with real testbench data.

Applications & Benefits

Are there demonstrators?

Yes, OSxCAR develops several demonstrators:

• Single-Vehicle: Basic SDV functionality on a single vehicle
• Multi-Vehicle: Scaling to more than 10 nodes for V2X scenarios
• Including benchmarking and test framework integration

Which industries is OSxCAR relevant for?

While focusing on automotive, the architecture is transferable to:

• Robotics: Autonomous robot systems
• Avionics: Aircraft electronics
• Medical technology: Networked medical devices
• Cloud infrastructures: Edge computing systems

What concrete benefits does OSxCAR offer?

• Cost reduction: Up to 50% savings through software reuse
• Faster time-to-market: Accelerated development cycles
• CO₂ reduction: Up to 20% through AI optimization and efficient software
• Quality improvement: Structured testing with clear interfaces (WASI/WIT)
• Scalability: From MIL/SIL via HIL/VIL to production with one binary

Standards & Compliance

Which standards are supported?

OSxCAR is based on open standards:

• W3C WebAssembly: Standardized runtime and binary format
• WASI (WebAssembly System Interface): System interface standard
• COVESA VSS: Vehicle Signal Specification for standardized vehicle APIs
• AUTOSAR Adaptive: Integration with automotive software architecture
• ISO 26262: Functional safety (ASIL qualification)
• ISO 21434: Cybersecurity for vehicles

Is OSxCAR ASIL-qualified?

The WebAssembly sandboxes are designed for ASIL qualification and meet the requirements of ISO 21434 and ISO 26262. Qualification occurs incrementally during the project duration.

Security & Privacy

How is sensitive information handled?

Non-public content (e.g., internal links, credentials, proprietary algorithms) is not published. The testbench infrastructure is TISAX-compliant for secure data processing.

How secure is WebAssembly in vehicles?

WebAssembly offers multi-layered security:

• Memory Safety: No buffer overflows or memory errors
• Sandbox isolation: Apps can only access explicitly released resources
• Capability-based security: Fine-grained permission control
• Deterministic execution: Predictable behavior for safety-critical systems

Next: Contact • Glossary • Downloads • Project Overview