Unlocking the Future of Software-Defined Vehicles with SOAFEE: A Cognizant Blueprint

The automotive industry is undergoing its most profound transformation since the invention of the car. Vehicles are evolving from hardware-centric machines to software-defined platforms capable of continuous updates, personalized services, and intelligent decision-making at the edge. Yet, realizing this vision is not without challenges. 

Modern vehicles can contain 100+ Electric Control Units (ECU), and bringing new features to market often takes 2–3 years. Customers are now demanding something very different: 

• Configurable, on-demand features that can be updated like smartphone apps.
• Seamless lifetime value delivered through software upgrades and subscription services.
• Real-time intelligence, leveraging AI to improve safety, efficiency, and customer experience. 

However, rising software complexity, tight coupling of hardware and software, and high validation costs are slowing innovation. The path forward demands a paradigm shift: from hardware-centric engineering to agile, software-first development, supported by scalable platforms, continuous integration/continuous deployment (CI/CD), and robust cybersecurity frameworks.  

Key Enablers: Building Blocks for Software-Defined Vehicles 

Cognizant’s SOAFEE Blueprint is rooted in four foundational enablers: 

• Cloud-native paradigm – Design, test, and deploy in the cloud, then seamlessly run in-vehicle.
• Cloud-to-edge architecture – Distribute workloads between the vehicle and the cloud dynamically.
• Automotive DevOps – Manage development, integration, and operations with continuous delivery pipelines.
• Software portability – Enable “write once, run everywhere” through service-oriented architecture. 

Cognizant has engineered a modular, service-oriented Software-Defined Vehicle (SDV) platform that integrates SOAFEE (Scalable Open Architecture for Embedded Edge) principles with advanced AI-driven analytics. Spearheaded by Arm and backed by a diverse ecosystem including Cognizant, SOAFEE offers a standardized, scalable framework for building, deploying, and orchestrating SDV applications across heterogeneous hardware environments, from embedded edge devices to cloud-native infrastructure. 

• Cognizant’s SOAFEE Blueprint enables OEMs and Tier-1 suppliers to accelerate innovation, reduce time-to-market, and deliver dynamic, software-centric vehicle experiences. Key components include Digital Cockpit SOAFEE Image – Yocto-based SOAFEE EWAOL (Edge Workload Abstraction and Orchestration Layer) builds, containerized In-Vehicle Infotainment (IVI), and cluster apps using AWS EC2 instances and automation via CI/CD pipelines.
• Virtual Vehicle Validation – Cloud-hosted Human Machine Interface (HMI) and Digital Cluster simulation using Functional Mockup Units (FMUs).
• AI-Driven Tire Safety Monitoring – Real-time Tire Pressure Monitoring System (TPMS) sensor data analyzed by LLM-powered inference engines (Meta’s LLaAM3, Alibaba’s Qwen2.5) for predictive maintenance, reduced fuel consumption, and driver alerts.
• Cloud-to-Edge Consistency – Bit-parity validation across NXP i.MX8 hardware and AWS cloud instances ensure portability and trust. 

Figure 2 illustrates Cognizant’s SOAFEE Blueprint with integrated AI capabilities, deployed across three distinct Amazon EC2 instances: 

1. Digital Cockpit (comprising Cluster and IVI subsystems)
2. AI-Driven Monitoring Engine
3. Elektrobit (EB) Corbos Adaptive Core 

The Digital Cluster and IVI middleware are modeled and simulated using MathWorks®  MATLAB® Simulink, where functionalities are validated through Model-In-the-Loop (MIL) testing. Source code is auto-generated using Embedded Coder, configured to enforce MISRA C compliance, ensuring that all generated code adheres to industry-standard safety and reliability guidelines for automotive software. 

Subsequent Software-in-the-Loop (SIL) testing is performed using dSPACE SystemDesk and VEOS, enabling early verification of runtime behavior and interface integrity. Once validated, the middleware components are containerized and deployed independently on AWS Graviton4 instances, running Yocto Linux with the EWAOL (Edge Workload Abstraction and Orchestration Layer). This deployment ensures bit-parity between cloud and edge environments, supports modular rollout, and aligns with SOAFEE’s cloud-native principles, delivering scalable, secure, and standards-compliant software-defined vehicle functionality. 

At the heart of the TPMS analytics pipeline is a Service-Oriented Subscriber, architected to process real-time telemetry from the TPMS Sensor Service Publisher as shown in figure 3. The publisher continuously streams granular data including tire pressure, temperature, and wear metrics from each wheel. This data is transmitted using a Service-Oriented Architecture (SOA), which ensures modularity, scalability, and loose coupling between components, allowing services to evolve independently and be deployed across heterogeneous platforms. 

To maintain compatibility with legacy automotive systems, the architecture also supports traditional signal-based communication protocols, such as CAN (Controller Area Network) and LIN (Local Interconnect Network). These protocols are used to transmit raw sensor signals from the TPMS hardware to the edge gateway, where they are abstracted into service messages. A middleware translation layer converts these low-level signals into structured service payloads, enabling seamless integration with modern cloud-native components. 

The subscriber service is designed to handle high-frequency data streams using the Google Remote Procedure Call (gRPC) open-source framework for low-latency, and bidirectional communication across different environments. It supports event-driven processing, where each incoming message triggers real-time analytics, anomaly detection, and feature extraction. This hybrid approach of combining SOA with signal-based protocol ensures backward compatibility with existing vehicle networks while enabling forward-looking, scalable software-defined vehicle architectures. 

The Blueprint integrates large language models (LLMs) and PyTorch-based inference for dynamic, prompt-driven analytics. By leveraging Elektrobit Adaptive AUTOSAR for standardized data exchange, the solution delivers: 

• Real-time monitoring of tire health.
• Predictive analytics to reduce downtime.
• Insightful dashboards for maintenance and fuel optimization.
• Scalable, CPU-optimized AI deployment on Arm-based cloud, avoiding GPU cost overhead. 

The Tire Pressure Monitoring System (TPMS) AI-Powered Data Analytics pipeline as shown in figure 4 is built on a service-oriented architecture, enabling real-time, modular, and scalable processing of sensor data from each wheel. Tire pressure, temperature, and wear metrics are streamed via a publisher-subscriber model such as SOME/IP, ensuring low-latency and reliable communication between components. gRPC acts as an inter-process communication subscriber side to pass the receive events from Publisher to the AI Analytics Engine. 

Incoming data is ingested and preprocessed with Python pandas, allowing for efficient cleaning, transformation, and structuring into time-series formats. This prepares the data for both visualization—using Matplotlib to highlight trends and anomalies—and advanced analytics. 

At the core of the analytics engine, large language models (LLMs) such as Qwen and LLaMA3 are deployed on Arm Neoverse V2-based AWS Graviton4 CPUs, optimized with Arm KleidiAI for high-throughput, energy-efficient inference. These models, integrated via Hugging Face Transformers and PyTorch vLLM, use few-shot prompting to interpret sensor anomalies and generate actionable diagnostics in natural language. 

The result is an intelligent, scalable, and cost-effective TPMS solution that delivers real-time insights, predictive maintenance alerts, and comprehensive reports—fully optimized for Arm-based cloud infrastructure. 

Why Choose Cognizant’s SOAFEE Blueprint? 

For automotive OEMs and Tier 1 Suppliers, our key customers, navigating the shift to software-defined vehicles, Cognizant’s SOAFEE Blueprint offers more than a technical framework, it’s a strategic enabler for agility, scalability, and long-term competitiveness. Here’s how customers can leverage the blueprint to unlock real business value: 

• Accelerated Hardware Abstraction & Platform Scalability 

The Blueprint decouples software from hardware, allowing developers to deploy applications across multiple platforms that: 

• Reduces dependency on hardware
• Enables faster scaling across vehicle models and variants
• Achieves early defect detection
• Modular Software Delivery via Containerization 

The Blueprint is built using scalable containerized workloads and microservices empowering customers to: 

• Deliver configurable, on-demand features like enhanced Infotainment or ADAS upgrades
• Isolate safety critical functions from non-critical apps for improved fault tolerance
• Enable rapid software updates and feature rollouts through CI/CD pipelines
• Early Validation Though Simulation & Digital Twins 

The Blueprint provides a digital twin environment that allows developers to simulate vehicle behavior before hardware is available. This leads to: 

• Faster prototyping and reduced physical testing cost
• Continuous validation throughout the software development lifecycle
• Improved software quality and regulatory compliance
• Enables “Shift-Left” methodology
• AI-Ready Infrastructure for Intelligent Features 

The Blueprint provides scalable architecture that supports edge AI inference which allows customers to deploy real-time intelligence across vehicle domains. This enables: 

• Predictive maintenance using sensor analytics
• Personalized driver experiences based on behavioral modeling
• Enhanced safety through autonomous decision-making at the edge. 

By adopting Cognizant’s SOAFEE Blueprint, OEMs and Tier 1 Suppliers gain a future-proof architecture for building intelligent, adaptable, and secure vehicles, while transforming their development processes to match the pace of digital innovation. This holistic approach not only speeds up innovation but also ensures that every software release meets the highest standards for safety, reliability, and user experience, empowering our customers to deliver cutting-edge, compliant automotive solutions with confidence. 

Conclusion 

Achieving truly software-defined vehicles demands more than adopting new technologies. It requires an open, scalable, and AI-enabled framework that can adapt to rapid industry change. Cognizant’s Blueprint, grounded in SOAFEE principles, illustrates how automotive organizations can accelerate innovation cycles, enhance vehicle safety, and deliver richer, more personalized customer experiences by embracing modular architecture and intelligent analytics. This positions the automotive industry to move beyond traditional limitations, making software the driving force behind future mobility. With SOAFEE, the vision of agile, intelligent, and customer-centric vehicles is not just a possibility, it is already becoming a reality. 

Ready to Revolutionize Your Vehicle Software Strategy with AI Intelligence? 

Get in touch today: [email protected]