This week we introduce next AIthena consortium partner: TTTech Auto
TTTech Auto provides solutions for future vehicle generations and the shift towards software-defined vehicles. The company specializes in delivering safe software and hardware solutions along with a range of services, to support the advancement of automated driving and beyond, applicable in series production programs. With its leading technology solutions, TTTech Auto reduces the required time for system designing, integration and testing of a series production ready software platform. This ensures safety and electronic robustness for a more automated world.
TTTech Auto was founded in 2018 by TTTech Group and technology leaders Audi, Infineon, and Samsung to build a global, safe vehicle software platform for automated and autonomous driving. In 2022, the company raised USD 285 million (EUR 250 million) from Aptiv and Audi in its latest funding round. At TTTech Auto’s headquarters in Vienna, Austria, and in more than 10 locations across Europe and Asia, 1,100 employees work with leading car manufacturers on their software-defined vehicle, ADAS and autonomous driving programs.
What is the role of TTTech Auto and the team in the project?
The TTTech team focuses in AIthena on further research towards software-defined architectures for the ADAS (Advanced Driver Assistance Systems) and AD software components to orchestrate services, guarantying safety, and cybersecurity, enabling over-the-air updates. The R&D activities in the project go towards safety-by-design by exploring deterministic virtualization platforms utilizing full system virtualization as well as containerization. Aspects of deterministic virtualization, such as scheduled virtual networks to enable predictable communication of virtualized components, is relevant for functional safety applications, e.g., emergency electronic breaking, speed harmonization, synchronizing and coordinating trajectories, or collision avoidance. The resulting virtualized real-time runtime will be deployed in the “Robust Prediction modules for Robo-taxi in urban environment” demonstrator. The real-time runtime prototype that will be developed within this Horizon Europe funded cooperative project can act as a step towards enabling customers and automotive OEMs worldwide to access innovative solutions in the future.
What are you currently working on in the project?
TTTech Auto is currently working on the finalization of the deterministic virtualization concept prototype utilizing time-triggered full system virtualization and time-triggered containerization for components with stringent timing requirements. Additionally, we are finalizing demonstrator requirements in cooperation with our partners Virtual Vehicle, Infineon, Siemens, and Continental. The partners contribute components targeting AI-supported perception and decision-making algorithms executing on top of the virtualized real-time runtime. In the following months TTTech Auto will complete the reference implementations for a preliminary integration in the second year of AIthena. By these contributions, TTTech advances the state-of-the-art setting the project up to enable the applicability of obtained results for next level of automation supporting AI-enabled CCAM (Connected, Cooperative and Automated Mobility).
From your perspective, how do you see the contribution of the AIthena project to building trustworthy, explainable, and accountable AI-based CCAM?
The basis for a trustworthy, explainable, and accountable AI-based CCAM is a deterministic runtime environment that enables a timely execution of components to ensure safety of a highly automated vehicle. A lose coupling of components, as enabled by the time-triggered paradigm, allows for independent component-based tests ensuring their correct functioning (explainable and accountable). The prototype of virtualized compute and communication architecture developed by TTTech Auto featuring deterministic full system virtualization and containerization ensures spatial and temporal isolation of loosely coupled components and improves trust in the system. Furthermore, containerization allows for fast recovery times of components in case of failure (availability) supplementing full system virtualization enabling low-cost redundant architectures easing the implementation of fault-tolerant services (reliability). The proposed mechanisms support addressing challenges in achieving functional robustness of the system.
Authors: Jan Ruh, Anna Ryabokon, Edin Arnautovic
You can read more about TTTech Auto at Leading software and hardware platforms from TTTech Auto (tttech-auto.com), and TTTech Group at Technologies that improve human lives – TTTech