Autonomous driving and electrification are key innovation drivers in the automotive industry. They rely on efficient Edge AI solutions to support essential functions like driver behavior monitoring, pedestrian recognition, traffic sign identification, and trajectory control.
Today, Imagimob, a subsidiary of Infineon Technologies AG, has strengthened its automotive machine learning portfolio by integrating AI capabilities into Infineon’s ASIL-D (Automotive Safety Integrity Level D) compliant automotive MCUs, such as the AURIX™ TC3x and AURIX™ TC4x.
Infineon’s AURIX TriCore microcontrollers combine a RISC processor core, microcontroller, and digital signal processor (DSP) into a single chip (MCU), making them ideal for complex automotive applications. These microcontrollers (MCUs), are designed for safety-critical systems in both traditional combustion engines and electric/hybrid vehicles.
The AURIX TC4x family is a next-generation evolution that provides enhanced performance, scalability, and secure processing for applications like electric mobility (eMobility), advanced driver assistance systems (ADAS), and automotive electronic/electrical (E/E) architectures. It integrates high-speed communication interfaces like 5G-bit Ethernet and new standards such as CAN-XL, offering flexibility for connected and autonomous vehicles.
The AURIX TC3x family, with its hexa-core architecture, supports ASIL-D (Automotive Safety Integrity Level D) compliant systems and is commonly used for engine management, transmission control, braking systems, electric power steering, airbags, and sensor-based technologies like radar and LIDAR. It is particularly suited for safety-critical applications in autonomous driving and hybrid vehicles, including battery management and inverter control.
Both TC3x and TC4x families are key components in enabling secure, reliable, and efficient systems for autonomous, clean, and connected automotive technologies.
Imagimob specializes in enabling machine learning (ML) at the edge, particularly for autonomous and automated driving applications. They provide a platform called Imagimob Studio that allows developers to create, train, and deploy machine learning models directly onto microcontrollers (MCUs), such as Infineon’s AURIX TC3x and AURIX TC4x families. This simplifies the complex process of integrating AI capabilities into automotive systems.
Imagimob helps developers from automotive OEMs and Tier 1 suppliers build machine learning models tailored for edge computing, focusing on AI tasks like object detection, acoustic recognition (e.g., siren detection), and predictive maintenance.
The company integrates its ML platform with Infineon’s AURIX MCUs to allow automotive manufacturers to implement advanced AI applications that can operate in real-time and securely handle large amounts of data. Imagimob provides pre-built projects (e.g., siren detection) to help developers understand how to deploy ML models onto AURIX MCUs, reducing the learning curve and making the process more accessible.
Additionally, Imagimob has created new regression models capable of estimating remaining battery power, assessing health status, and calculating usage time.
The AURIX TC4x scalable MCU family is an upgrade of the TC3x series of ASIL-D compliant automotive microcontrollers, powered by the next-generation TriCore 1.8 architecture. It features a scalable accelerator suite with a parallel processing unit (PPU) and intelligent accelerators for cost-effective AI integration. These advancements enhance machine learning capabilities, allowing developers to simultaneously deploy multiple or more complex models. For example, while the TC3x delivers basic siren detection, the TC4x can manage both siren detection and voice interaction at the same time.
We have covered a few strategic partnerships between Infineon and software developers that enhance the capabilities of the AURIX TriCore microcontrollers, including the one with Aurora Labs and, more recently, Green Hills Software.