Develop the Functional Architecture for EV4.0 and EV5.0 and implement in ECS, enabled by Embedded Intelligence and Functional Virtualization for connected shared mobility.
The H2020-ECSEL/FFG project AI4CSM (“Automotive Intelligence for Connected Shared Mobility”) is motivated by the vision set out by the European Commission in the Green Deal. The project takes on the trends of Electrification, Standardisation, Automatization, and Digitalisation to achieve green and inclusive mobility with a focus on the automotive domain as shown in the figure below. Having assembled some of Europe’s finest from industry, research and academic, AI4CSM will deliver key innovations in technical areas these trends are rooted in: (1) Sensor fusion, perception systems and advanced integration platforms, (2) Efficient propulsion and energy modules, (3) New wired and wireless connectivity and architectures for cooperative mobility applications, (4) New Vehicle/Edge/Cloud computing integration concepts, (5) Digital platforms for efficient and federated computing, and (6) Intelligent components based on trustworthy AI techniques and methods. Notice that “standardisation” in this context refers to standardised components and common interfaces.
For example, AI4CSM’s intelligent in-car platforms will cut down on the number of components in a car, provide a virtualised execution environment with the computational power needed to run trustworthy AI algorithms, and make flexible on-board/off-board computation available as the platforms will include predictable latency and high bandwidth connections, making them fully connected and cooperative architectures. Currently, such platforms are not available. Besides intelligent platforms, future automated cars critically need dependable perception and intelligent components, AI4CSM will advance these areas. Basis to being able to add more intelligence in a safety critical domain such as mobility is trustworthy AI and the verification and monitoring of AI-based algorithms, which AI4CSM will cover as well. In completing the systematic approach, AI4CSM will improve green propulsion aiming at more intelligent control and a further reduction of energy consumption through improved components.
Fakten
Project duration: May 2021 – April 2024
Coordination: Infineon Germany
Budget: € 41,8 Million
Funding: EU: € 11,9 Million, national: € 11,7 Million
Partner:
- Industry: AVL (AT), Infineon Austria (AT), Mercedes-Benz (DE), NXP (DE,NL), SINTEF (NO), TTTech (AT)
- SME: AI Digital Solutions (DE), Ideas & Motion (IT), Innatera Nanosystems (NL), UAB Teraglobus (LT)
- Academia: Politecnico di Torino (IT), TU Brno (CS), TU Delft (NL), TU Graz (AT), TU Vienna (AT), ViF (AT)
