KI-Absicherung positions Germany for the new key technologies and market leadership of the German automotive industry with regard to automated driving.
We were part of the „KI Absicherung“ research project that is part of the Federal Government’s AI strategy, which positions Germany in the long term for the new key technologies and market leadership of the German automotive industry with regard to automated driving with an overall budget of 41 million euro.
In these initiatives, we have focused on the Brittleness safety concerns. In this aspect of the work, we are able to use our experience to develop test methods for the task of reliable detection of pedestrians on the street. neurocat developed test methods and integrated them into our proposed innovative robustness test strategy and provided guidance on how to assess the risks associated with deploying this complex computer vision task in an autonomous vehicle. Furthermore, we helped the developers of the Deep Neural Networks to mitigate risk by providing recommendations on where to improve the perception algorithms in line with the safety goals of the project.
An autonomous vehicle must perceive its environment and react appropriately to it. This environmental detection must be able to detect the movements of others in order to interpret road users, like pedestrians.In highly automated vehicles, these tasks will increasingly be adopted by artificial intelligence (AI), as AI functionality modules based on machine learning are becoming a key technology.
One of the biggest challenges in integrating these technologies into highly automated driving cars are the established processes of functional safety that cannot be without further ado transferred to machine learning methods.
To solve this challenge, the research project „KI-Absicherung“ was launched on 1 July 2019 consisting of 25 partner companies and institutions led by Volkswagen AG. The aim of the consortium is to achieve a stringent and verifiable argumentation chain for the validation and deployment of AI function modules in context of highly automated driving.
In the „KI-Absicherung“ project, leading experts from industry and science work out so far largely independently acting disciplines of AI algorithms, the 3D visualization and animation as well as functional safety together for the first time. Over the last three years, we worked together to develop solutions to monitoring, evaluating, testing and correspondingly reliable and to be able to secure transparently.
We worked together with 25 Partners:
- Automotive Manufacturer: Volkswagen AG (Konsortialführer), AUDI AG, BMW Group, Opel Automobile GmbH
- Suppliers: Continental Automotive GmbH, Hella Aglaia Mobile Vision GmbH, Robert Bosch GmbH, Valeo Schalter und Sensoren GmbH, Visteon Electronics Germany GmbH, ZF Friedrichshafen AG
- Technology Provider: AID Autonomous Intelligent Driving GmbH, Automotive Safety Technologies GmbH, Intel Deutschland GmbH, Mackevision Medien Design GmbH, Merantix AG, Luxoft GmbH, umlaut systems GmbH, QualityMinds GmbH
- Research Partners: Fraunhofer IAIS (Stellv. Konsortialführer und Wissenschaftlicher Koordinator), Bergische Universität Wuppertal, Deutsches Forschungszentrum für Künstliche Intelligenz, Deutsches Zentrum für Luft- und Raumfahrt, FZI Forschungszentrum Informatik, TU München, Universität Heidelberg
- External Technology Partners: BIT Technology Solutions GmbH, neurocat GmbH, understand ai GmbH
- Projektmanagement: European Center for Information and Communication Technologies – EICT GmbH
More information can be found here: https://www.ki-absicherung-projekt.de/fileadmin/KI_Absicherung/Downloads/KI-A_20200221_Presseinfo_v1.1.pdf