PhD Student Distributed Generative AI for Robotics

The Fraunhofer-Gesellschaft currently operates 76 institutes and research institutions in Germany and is one of the leading organisations for application-oriented research. Around 32 000 employees develop the annual research volume of 3.4 billion euros. The newly founded Center for Applied Artificial Intelligence Duisburg (ZaKI.D) aims to support the regional economy in the transformation in Duisburg and in the adjacent regions with promising AI technologies. The technical focus of the ZaKI.D is the embedded artificial intelligence on resource-restricted devices such as sensors, smart devices or robots. We want to make these safer and smarter, optimize manufacturing processes, increase the quality of products and minimize the ecological footprint. Distributed Generative Artificial Intelligence (AI) is an innovative approach in which the principle of Federated Learning is used to accelerate parallel learning of robotic skills. This method enables decentralized data processing, in which AI models are trained directly on the devices or in a large number of simulations, without the need to centrally store sensitive data. As a result, the autonomous action capability of robots is improved and their interaction with the environment is optimized. The aim of the doctoral thesis is to investigate the underlying methodology and algorithms, to develop new methods and their analysis and assessment. The developed methods are to be evaluated by implementing a framework for distributed generic AI training, which enables the integration and cooperation of various robots in a network, within the framework of concrete implementation projects with customers in the ZaKI.D.

First of all, look at the state of science and develop the theoretical foundations of the Federated Learning, the distributed generative AI and the Reinforcement Learning. Based on existing libraries and frameworks (e.g. ROS, ROS2), you will establish a simulation environment for distributed model training and evaluation processes. Using concrete questions in the field of production with robots, you develop, implement and evaluate new methods in the field of distributed artificial intelligence (e.g. for trajectory generation, robotic manipulation). In close cooperation with other employees in the ZaKI.D, you implement the methods in a distributed generic AI framework that you use in practice within the framework of ZaKI.D implementation projects for industrial customers. You will publish the results of your experimental work as scientific publications and present them at international conferences and in the project meetings with external partners. In addition, you will supervise student and scientific assistants as well as bachelor and master thesis.

Highly completed studies (M.Sc. or Uni-Diplom) in computer science, mathematics, physics, electrical engineering, mechanical engineering or a comparable technical/nature science course In-depth know-how (ideally in the form of a master's or diploma thesis) in the field of artificial intelligence with focus on reinforcement learning, generic AI or in the field of control technology for robotics Knowledge in the programming language Python and with relevant tools/frameworks such as PyTorch, Scikit-Learn, ROS/ROS2 or the like. A systematic and independent way of working, coupled with creativity, hands-on mentality and communication security in German and/or English

Doctorate in an application-oriented topic within 3 years. You can access all the institute's own clean rooms and laboratories for your practical work. Teaching obligations will not be transferred to you. The doctoral degree is awarded by the University of Duisburg-Essen. During the promotion period, you will be supported by accompanying offers. For example, you can benefit from regular doctoral student coaching, additional support from your institution-internal caregiver as well as the half-yearly doctoral lecture before our doctorate/patient body with direct feedback round. In addition, you will learn to apply professional methods of project management safely during your doctorate and to support the project acquisition. If the dissertation is submitted in due time, the option is to further deepen the research work within the framework of a connection agreement or to switch to other areas. The full-time post as a doctoral student with half-remuneration offers 50% of the time for your doctoral degree and 50% for cooperation in research projects as a scientific assistant. Flexible working hours (guiding time with integrated core working time of basically 9:30–15:00, Friday 9:30–13:00) and mobile work on up to two days for better reconciliation of work and private life Occupational pension (VBL) and grant to the Deutschland-Ticket Job Very good transport with public transport/car as well as free parking and covered bicycle parking for employees Family and Occupational Compatibility Support: Mit-Kind-Office, Kindernotpflege und Beratungs zu Homecare-Eldercare etc. in cooperation with the pme family service Corporate Benefits: Benefits Offers from renowned manufacturers and brands JOBV1_EN