Embedded Software Development Practices

When developing real-time embedded software for our radar sensors, we prioritize quality, safety, and sustainability. Given the safety-critical nature of these systems, both static and dynamic code analyses are performed continuously using advanced toolsets throughout the development lifecycle.

Code implementation is managed with Git-based version control systems. Code readability and maintainability are enhanced through structured code-review processes, while systematic unit and integration testing ensure the software functions reliably across different layers.

Code Validation and System Monitoring

To improve code quality, we use a combination of static analysis tools and dynamic analysis techniques—such as runtime monitoring and memory-leak detection. Based on these findings, we refactor the code to enhance its resilience and reduce the risk of defects.

In parallel, we monitor key system-level performance metrics including stack and heap memory usage, CPU load, and RTOS scheduler behavior. These continuous insights guide us in maintaining real-time system stability.

Compliance, Traceability, and Functional Safety

Our development process is traceability-driven: each software requirement is explicitly mapped to its implementation, streamlining validation and verification activities. CI/CD pipelines automate test execution and deployment at defined intervals, ensuring smooth and traceable software releases.

All code is compliant with MISRA-C guidelines and ISO 26262 Functional Safety standards. In the event of a fault, cross-functional teams—including hardware, RF, software, and mechanical engineers—collaborate within the ISO 26262 framework to implement both preventive and corrective safety mechanisms.

Advanced Algorithms and AI Integration

For radar-based target detection and tracking, we utilize advanced signal-processing algorithms that deliver precise range and velocity measurements—greatly enhancing environmental awareness in automotive scenarios.

ITAS | Intelligent Systems integrates AI into radar applications. We develop object-recognition and tracking algorithms for automotive radar and data analytics software for medical radar. In our in-cabin radar project, our machine-learning model achieved 92% accuracy in classifying occupants, meeting NCAP’s CPD (Child Presence Detection) criteria for automotive safety systems.