RM scientists in the finals for 2019 AI newcomer
With the project #KI50: Künstliche Intelligenz in Deutschland – gestern, heute, morgen (Artificial Intelligence in Germany – yesterday, today, tomorrow), the Gesellschaft für Informatik (GI) wants to spotlight dedicated young AI researchers from all disciplines. For this, the GI is honoring ten outstanding AI newcomers in the Federal Ministry of Education and Research’s Science Year 2019 – Artificial Intelligence.
Two scientists from the German Aerospace Center (DLR) have made it to the finals:
Annette Hagengruber and Xiangyu Zhuo
Annette Hagengruber studied Medical Engineering at the University of Applied Sciences Upper Austria and went to DLR for her master’s thesis. Here, she works in Assistive Robotics and in the group for adaptive bio-interfaces. The scientist mainly works with the robot assistant EDAN (EMG-controlled Daily AssistaNt) on the further development and implementation of partial autonomy, in which a user is supported by EDAN in everyday tasks such as eating and drinking. One essential aspect of this is that the user always has freedom of choice.
The scientist and her colleagues are also developing an interface that enables people with very limited mobility to regain the ability to control a robot assistant.
Xiangyu Zhuo has been working as a research associate at DLR since October 2014. In March 2019 she completed her doctorate at the Technical University of Munich, before that she studied photogrammetry at Wuhan University in China. Her research focuses on the identification of road markings and buildings in order to create a city map for applications in the fields of autonomous driving, traffic management and urban planning.
Xiangyu Zhuo says, “With my passion for AI, I am motivated to develop innovative solutions to real problems. One challenge for AI research is the lack of annotated training data. I developed a novel method to automatically generate image annotations that are as accurate as manual annotations. In a traffic and disaster-monitoring project, I use neural networks to extract semantic information from aircraft and unmanned aerial vehicle data. The fusion of multi-level information increases detailed knowledge of the scene, thus enabling fast responses in emergency situations in near real time.”
Annette Hagengruber says, “So far, I’ve tested EDAN’s capabilities only with people with spinal muscular atrophy. In another project, research will also be extended to nursing in general. For this purpose, we are planning studies on site in a nursing home. We want to develop useful applications for robotics and AI in the context of geriatric care. AI creates a multitude of new opportunities that we can use to make our assistance systems smarter, more intuitive and more robust for real-world environments. However, at this point AI must also be used properly. There’s a big difference between lab applications and those in the ‘real world.’ I’m working to unify AI and robots and turn them into real applications. I like to take on these challenges in order to develop useful and meaningful technologies.”