Infineon's Latest Automotive Microcontrollers Achieve Compliance with AI Safety Standards
At Embedded World 2024, Infineon introduced its latest automotive microcontroller (MCU) series featuring neural acceleration. With the growing influence of AI and machine learning across various sectors, the debate over the dependability and functionality of these intelligent systems is intensifying. Infineon's latest innovation aims to mitigate these doubts by equipping developers with MCUs that can safely integrate AI models into critical tasks, particularly in the automotive domain.
The AURIX TC4x MCUs. Image used courtesy of Infineon
Examining the AURIX TC4x Family: Advancing AI in Automotive MCUs with Enhanced Safety and Compliance
This in-depth analysis spotlights the AURIX TC4x series, the latest in automotive MCUs, and offers a comparison with its predecessors. The piece will also delve into the integrated AI safety features and compliance measures of the MCU series, which are intended to extend the reach of machine learning and AI-enhanced functionality across a broader array of applications.
Strengthening Trust in AI Systems
The AURIX TC4x series aligns with the SAFE AI principles established by the Fraunhofer Institute for Cognitive Systems IKS. This comprehensive safety framework is tailored to ascertain the reliability and security of AI models that operate high-stakes equipment, such as vehicles or industrial machinery. By adhering to these stringent safety guidelines, the TC4x family sets a precedence for incorporating AI in a broader spectrum of scenarios, some of which may have formerly been deemed unsuitable for automated control.
Juergen Schaefer, senior principal at Infineon's automotive division (left), discusses the importance and potential of SAFE AI compliance with All About Circuits editor-in-chief Jeff Child (right) at Embedded World 2024 in Nuremberg, Germany.
Enhancing AI Device Assessment with Infineon's SAFE AI-Compliant MCUs
Infineon's SAFE AI framework scrutinizes AI-driven devices on several fronts—robustness, data effectiveness, and operational design domain. To hasten regulatory compliance and market introduction, the TC4x MCU family has been fortified with built-in safety features. These provisions are designed to address both existing and potential future AI regulations.
While simpler AI models may not require such extensive measures, for intricate AI systems in charge of critical operations, conformity with the SAFE AI standards can significantly broaden the applicability of current AI methodologies.
Equipped for Multitasking Efficiency
Including a 500-MHz TriCore processor and a substantial 25 MB of integrated Flash memory, the TC4x series MCUs are ready for complex tasks. They also support connectivity with external memory solutions. The lineup is enhanced by the AURIX AI Accelerator Suite which features a dedicated parallel processing unit, a data handling engine, DSP core, radar accelerator, and security modules.
Collectively, this suite empowers the MCU to manage a multitude of automotive sensor applications, especially those involving radar technologies. With compatibility for prominent vehicle communication interfaces such as 5-Gbps Ethernet, PCIe, CAN-XL, and others, the chips are further optimized by the data routing engine to boost overall performance efficiencies.
The block diagram of the TC4x family highlights the AI acceleration and safety standards built into the device. Image used courtesy of Infineon (Click to enlarge)
MCUs Meet Highest Safety Standards with Enhanced Security Measures
Adhering to ASIL-D and ISO 21434 standards, the MCUs ensure top-level safety and security capabilities. The addition of Safe DMA enhances the intrinsic security of the new MCU family. Infineon streamlines the development process with comprehensive SDKs and drivers for the TC4x series. The brand also maintains compatibility with legacy TC3xx algorithms and ecosystems to facilitate a smooth upgrade to the latest devices.
Revolutionizing Vehicle Dynamics through Advanced Safe AI
The advent of the SAFE AI framework and the TC4x MCU family heralds a potential shift in automotive technology. Imagine vehicles that can determine gear positions autonomously, eliminating the need for traditional position sensors.
While AI may not be the universal remedy, the heightened safety and security measures embedded in the TC4x series could broaden the implementation of machine learning and artificial intelligence across a vast array of critical functions. This is particularly true as the confidence and dependency of users and developers on AI technology grow.
