Cross Section Cluster

The subject areas combined in the Cross Section Cluster make a significant contribution to the realisation and utilization of the three application clusters already introduced.

 

AI methods offer the possibility to capture complex nonlinear relationships in data and to develop algorithms that are superior to conventional model-based approaches in many fields of application. They form the backbone to make the complexity of autonomous mobility and the challenges of future automobile production manageable.

AI algorithms can already develop their potential in the perception phase, i.e. at the beginning of the signal processing chain of many applications, namely in the extraction of relevant information from sensor signals. The extraction of relevant information from sensor data is closely linked to AI methods for time series analysis and the subject area known in AI as representation learning. AI-based prediction and planning tasks build on the results of perception. They play a key role in autonomous mobility applications for predicting the behaviour of other road users and for route or trajectory planning.

In order to put AI methods into practice in a wide range of applications, especially in safety-critical applications of autonomous mobility, the interpretability and traceability of these methods is required for reasons of liability and approval, but also for ethical considerations and reasons of social acceptance. Further challenges that are counted among the topic "AI methods" include real-time capability and thus the question of suitable architectures of AI procedures as well as the efficient management of large amounts of data.

An important component of the AI mobility node is the intelligent infrastructure. It provides the basis for the development and testing of autonomous mobility. In Ingolstadt, there is a basic equipment that can be expanded with the "First Mile", the intelligent traffic light system and the test field for 5G. Together with companies and municipal partners, THI is planning a high-definition test field in which elements of the infrastructure, such as street lamps, will be equipped with sensors and generate information for mobility applications. This information is to be processed locally using AI methods and transmitted to a mobility centre via the 5G communication standard. In addition, the CARISSMA test sites offer the possibility to conduct road tests for critical traffic scenarios.

In addition to the mobility node, the end points in Regensburg, Aschaffenburg and Landshut also have test fields in the AI Mobility Network, which are designed for the development of vehicle automation. 

The topic "AI business models and services" focuses on the research of new mobility concepts and novel, partly disruptive mobility business models in which data and AI procedures are an elementary part of the business model. Particular attention is paid to intermodular mobility concepts, in which automotive mobility is linked to public transport, individual micro-mobility and urban air mobility. In addition, the subcluster will work out the factors for acceptance and success of these new business models from the user and provider perspective.

It is to be expected that within the framework of the AI mobility node, innovative mobility concepts will be realised prototypically as use cases and lighthouse projects. The consistent consideration of AI-based business models in the development of prototypes is expected to lead to the establishment of flagship projects with established companies. The work in this topic area should also contribute to the further development of OEMs into mobility service providers, for example by researching AI applications in the field of smart services to increase comfort when driving or flying.

Trust and acceptance of new technologies are a basic prerequisite for the success of innovative mobility applications. In the subject area "Ethics - Acceptance - Technology Impacts", the following aspects in particular are to be made usable for technology development, taking into account the requirements for "Trustworthy AI" formulated by the EU Commission's High-Level Expert Group on AI: Mathematical foundations and models of human-compatible programming of AI algorithms; ethical conditions; social effects and questions of governance and regulation in the use of AI; causal relationships at the interface between humans and AI; citizen participation.

The aim is to actively involve citizens in the development, testing and introduction of AI-based autonomous mobility from the very beginning in order to promote social acceptance. In addition, complex questions concerning the acceptance of AI decisions are to be dealt with in interdisciplinary teams.

AI mobility professorships at the THI in the Cross Section Cluster

A total of five professorships are to be established at the THI in the cross section cluster with funds from the High-Tech Agenda Bayern; one of these will result from the AI competition Bayern 2020. An overview of the planned professorships is provided together with the professorships financed from own or foundation funds:

Contact

Scient. director AININ; Programme director and Acadmic Advisor "Automated Driving and Vehicle Systems" (Master)
Prof. Dr.-Ing. Michael Botsch
Phone: +49 841 9348-2721
Room: H024
E-Mail: