Professorships in the AI Mobility Node Ingolstadt / AININ
Under the roof of AININ, in addition to the professors already conducting research, a further 40 scientists with basic funding and, in the future, a corresponding number of third-party funded positions will develop AI mobility research at the THI and accompany the transfer into applications. The new AI mobility professorships are represented in all faculties of the THI and cover the research fields of AI-supported automobile production (Faculty of Engineering and Management), autonomous driving (Faculty of Computer Science, Faculty of Electrical Engineering and Information Technology, Business School) and unmanned flying (Faculty of Computer Science, Faculty of Mechanical Engineering, Business School). The new AI research professorships are shown in the diagram:
High-tech agenda Professorships for the AI Node
The aim of the professorship is to research the advantages of AI processes in digital production and to bring them to application. One focus is the development and use of AI methods for quality assessment in production processes, so that quality assurance steps can be reduced or eliminated. By means of online feedback of correction parameters derived from the AI models, a constant product quality with greatly reduced scrap is to be guaranteed.
The professorship is to be closely interlinked with the AUDI endowed professorship and with a planned institute in the Industry 4.0 topic area. In this competence field, 8 professorships are currently conducting research on topics relevant to AI applications, such as networking and cloud architectures in production, information models and semantics, and additive manufacturing technologies.
A further topic area that is considered in this application cluster is the interaction of people and learning systems in production processes.
In addition to cloud-based services, edge computing (calculations and analyses at the very ends of an IT topology) is playing an increasingly important role in automated and networked driving. All computing operations of a networked mobility system that take place in a vehicle can be assigned to edge computing.
The aim of the professorship is to research new methods and procedures at the interface between hardware and software for the use of AI methods in edge computing for automated driving. The focus is on the consideration of the available computing resources.
The aim of the professorship is to use AI-based methods to generate a comprehensive image of the vehicle environment in real time from different sensors as a basis for automated driving. A robust sensor data fusion, which considers sensor data of different origin (e.g. from the vehicle or from the infrastructure) and different quality, is the focus of the research. Deep knowledge in the field of optical sensors (lidar, kamara) and radar sensors is necessary for this professorship.
The software for autonomous mobility systems is highly complex due to the high degree of networking of the large number of sensors and control units. The goal of the professorship is to extend established software architectures for automated driving in order to efficiently integrate AI algorithms and thereby meet the fail-operational requirements for autonomous systems.
AI can also be used to build intelligent and adaptive software systems and to support more efficient software development processes. Especially in the field of testing software for autonomous mobility it is expected that AI methods can make an important contribution. The professorship will also focus its research on these aspects of software development.
The aim of the professorship is to work on new techniques for the management and analysis of mobility data and to gain insights into mobility ecosystems using unsupervised learning methods. The efficient access to huge amounts of data is of great importance in the development as well as in the protection of automated vehicles, especially in connection with the use of AI methods. The automated identification of relevant traffic scenarios by means of AI methods and the associated possibilities of time-lapse are very important for the further development and testing of sensors and planning algorithms. The efficient data management in mobility centres forms the basis for smart services.
The real-time capability and the development of AI systems for mobility applications that do not require a lot of computing power are the focus of this professorship. Many system components required for automated mobility require real-time capability, e.g. detection and tracking of objects or control systems. In order to make the advantages of AI methods available for such system components, it is necessary to process AI algorithms in real-time.
Innovative architectures of AI algorithms and their efficient implementation on hardware components play a central role for the professorship. Thereby, profound knowledge in the field of software parallelisation for multi-core processors is required.
The aim of the professorship is to conduct research on methods to enable the systematic use of AI methods in aeronautical engineering. The focus is on the field of "Machine Learning Control", i.e. the use of AI for the design and operation of control systems in unmanned aircraft. In manned aviation, depending on the aircraft, decisions have to be made permanently in order to operate the aircraft optimally. These decisions are to be mapped by means of artificial intelligence in order to support unmanned aviation. In the development of mobility products in particular, tests and calculations take place regularly. The use of simulation tools for calculations has already led to an optimisation of efficiency in product development. Simulation calculations and also cost calculations can indeed be based on similarities. However, in many cases the estimates required for this are still dependent on human beings. It is to be investigated how similarity considerations can be represented by means of AI and actively used in product development in order to be able to carry out product developments more quickly and with less expenditure of resources.
The aim of the professorship is to use AI methods to work on solutions to the challenges of flight guidance for unmanned flight. The application focuses on flight management systems with increasing autonomy for classical aircraft as well as UAM. The combination of current AI methods and the research of machine learning procedures for safety-critical systems should enable efficient solutions for the following applications: Processing results of environment acquisition and replanning of the trajectory, flight guidance for fully autonomous flying for optional manned and unmanned flight systems incl. UAM as well as the integration of unmanned flight systems into the airspace.
The use of AI requires social acceptance. Therefore, the societal implications in terms of technology assessment of AI applications have to be investigated. In this context, an ethical consideration is also necessary. The lack of transparency and explanation of automatic decisions or recommendations made by AI algorithms is critical. This societal and ethical view on the consequences of the use of AI in turn results in requirements for the development and testing of mobility systems and smart services based on AI procedures. The aim of the professorship is to conduct research at the interface between technical and societal implications of the use of AI in mobility systems and to develop basic principles for the social acceptance of AI mobility applications.
The aim of the professorship is to work on methods to make passenger and freight transport more efficient, sustainable and comfortable. The focus is on AI-supported optimisation and interlocking of classic mobility offers (e.g. public transport with individual transport), individual mobility concepts (e.g. sharing offers) and parking and energy supply infrastructure (e.g. charging stations). In particular, smart services for urban areas should be considered within the framework of increasing efficiency, sustainability and comfort. The focus will be on the economic, ecological and social aspects of the use of smart services. The research area will combine macroscopic and microscopic traffic observations and investigate the benefits of mobility applications also from a business perspective.
The professorship is designed as a central cross-sectional professorship in the AI mobility network. It will research innovative mobility concepts that are made possible through the use of AI methods with the aim of making a positive ecological, economic and social contribution. The professorship will focus on the research of sustainable business models in the field of AI-based mobility and their application by start-ups or established companies (intrapreneurs).
The professorship will assume a comprehensive coordination role by researching the applicability of research results in the entire mobility network. It is networked with all other research professorships of the node and endpoints. Due to its innovation and organisational relevance, the professorship is to assume the necessary coordination role in the network of the AI node and the AI endpoints, for example in the development of cross-university use cases and flagship projects or the organisation of subject-specific scientific exchange formats.
Professorships from foundation or own funds
The endowed chair supports the goals of the application cluster "AI-supported automotive production" and focuses on the optimization of logistics and production processes through AI-supported analyses and forecasts.
The endowed chair of the city of Ingolstadt „Nachhaltige Stadtentwicklung und Künstliche Intelligenz“ addresses on the one hand prerequisites in the infrastructure, which are necessary for automated driving and autonomous flying, and on the other hand the macroscopic view of traffic, with the aim of optimizing the traffic flow by means of data-based procedures. Another focus of the professorship is the research of AI methods for mobility and infrastructure planning in urban areas as fundamental tools for the sustainable design of urban living space.
The self-financed professorship for "Text and Speech Comprehension" is intended to contribute to an optimized human-machine interface. The research work focuses on AI methods for processing time series such as recurrent neural networks, but also on new possibilities for semantic description of language. Architectures of AI speech processing systems for speech recognition and speech modelling will be investigated. It is to be expected that in the future a substantial part of the communication between humans and machines will take place via speech. Thus, this professorship also provides central interfaces to the three application clusters of the AI mobility node.
Job holder: Prof. Dr. Munir Georges (start of employment on 01.09.2020)
The self-financed professorship "Computer Vision for Intelligent Mobility Systems" focuses on advanced AI methods in the field of image processing. Special attention is paid to boundary conditions that have to be considered when using cameras in mobility vehicles, such as adverse environmental conditions or high dynamics. The aim is to investigate AI architectures that can reliably guarantee a high detection quality from camera data even under such boundary conditions. Because cameras are among the most important sensors in the realization of automated driving, unmanned flight and condition monitoring in automotive production, this professorship has central interfaces to the researchers in the three application clusters of the AI mobility node.