Autonomous Driving and Edge AI: Imagimob and Infineon Team Up to Enhance Automotive Machine Learning

Autonomous and automated driving, along with electrification, is a key trend in the automotive industry. Artificial Intelligence (AI) plays a crucial role in this transformation, enabling vehicles to perform tasks such as detecting pedestrians, analyzing driver behavior, recognizing traffic signs, and controlling trajectories. Edge AI is a critical component in this trend, as autonomous driving demands AI systems with robust machine learning capabilities and processors that can handle vast amounts of data in real-time, safely and securely.

To meet this challenge, Imagimob, a company under Infineon Technologies AG, has strengthened its automotive machine learning portfolio by integrating machine learning (ML) capabilities into Infineon’s Automotive ASIL-D compliant microcontrollers (MCUs), such as the AURIX TC3x and AURIX TC4x families.

Thomas Boehm, Senior Vice President of Microcontrollers at Infineon, emphasized the importance of this development, stating,
“Integrating secure and reliable AI capabilities into microcontroller families is essential for the advancement of autonomous driving applications in the automotive sector. We are proud that our AURIX microcontrollers are now supported by Imagimob Studio, which makes them accessible to developers worldwide and solidifies our position as a leading innovator in the industry.”

Alexander Samuelsson, CTO of Imagimob, highlighted the platform’s versatility, noting,
“With the integration of AURIX into Imagimob Studio, we are bringing full machine learning capabilities to the automotive sector. This means all the use cases we support are now available for Infineon’s AURIX microcontrollers.”

With Imagimob Studio, developers can now create and deploy robust machine learning models for Edge AI on Infineon’s proven AURIX MCUs. The process starts with model creation in Imagimob Studio, after which users can easily deploy the model to the MCUs. The platform simplifies the entire process by providing step-by-step guidance, enabling developers to implement sophisticated ML models on MCUs with ease.

Additionally, Imagimob Studio includes a sample project for siren detection, showcasing the process of creating and deploying a model. This example allows users to explore how to build acoustic models using AURIX MCUs and a microphone shield. Imagimob has also developed regression models to calculate metrics like remaining battery power, health status, and usage time.

Advanced AI Use Cases with AURIX TC4x

The AURIX TC4x family of MCUs offers a seamless upgrade path from the AURIX TC3x series, providing enhanced performance thanks to its next-generation TriCore™ 1.8 architecture. The TC4x family also features a scalable accelerator suite, including a parallel processing unit (PPU) and multiple intelligent accelerators, enabling cost-effective AI integration.

These advancements in the AURIX TC4x series allow developers to deploy multiple or more complex machine learning models. For example, while the AURIX TC3x can manage basic tasks like siren detection, the AURIX TC4x can handle both siren detection and voice interaction simultaneously, showcasing its ability to support more advanced AI use cases.

Availability

Imagimob Studio is available at https://www.imagimob.com/studio, and further details about Infineon’s AURIX family of microcontrollers can be found at http://www.infineon.com/AURIX.