© amiak/Fotolia.com (modified)
In the Helmholtz Artificial Intelligence Cooperation Unit (HAICU) the Helmholtz Association of German Research Centres is setting up an innovative network for applied artificial intelligence (AI). HAICU will design, implement, and distribute AI methodologies for such purposes as analysing complex climate, energy, transportation and health systems. In this network the German Aerospace Center (DLR) in Oberpfaffenhofen near Munich was awarded the contract for the "HAICU Local" unit in the aviation, aerospace and transportation research fields. Artificial intelligence and machine learning are powerful tools for finding solutions to the challenges existing in these large and complex fields. DLR is a leader in the area of applied AI in earth observation and robotics, and is convinced of its high potential for addressing research questions connected with autonomous driving and aviation.
DLR will receive Helmholtz Association financing to establish a DLR "Junior Researchers Group" for AI in earth observation" as well as a "High Level Support Team". This team of experts will assist other institutes and Helmholtz centres with their AI projects and together with them implement joint projects.
AI in earth observation
"Globally available geoinformation derived from earth observation satellite data has become essential for many geo-related science, political, and planning issues" stresses Professor Xiaoxiang Zhu, who heads the Earth Observation Data Science department at the Remote Sensing Technology Institute as well as DLR's HAICU unit. Geosciences, environmental sciences, sustainable development, resource management, civil security, disaster control and decision support are prominent examples. In all of them AI is playing a continually increasing role since the associated methodologies are far superior to existing algorithms, especially for today's "big data" from earth observation. The demands for high quality geoinformation and the wide range of applications call for AI research and innovative methodologies specifically tailored to meet earth observation requirements.
AI in robotics
Safe human interaction is essential for many robotics applications in aerospace, disaster control, and personal care. On the other hand, applications like planetary exploration require a high level of autonomy. In any case, AI methodologies are very useful for achieving this safety, autonomy and reliability. Innovative machine learning procedures are being developed in the "Perception and Cognition" and "Cognitive Robotics" departments at DLR's Institute of Robotics and Mechatronics. Applications include teaching robots to rapidly and reliably identify and grasp objects, or the capability to carry out error analyses of previous collected data sets. Other important research at the institute relates to the persistence of learning methods in order to appropriately react to change, and to autonomous learning, or the ability to make decisions that improve the learning process without requiring significant human interaction.
As the third partner in DLR's HAICU unit, the Institute of Simulation and Software Technology is contributing its computing expertise in in order to optimize for big data and high performance computers the algorithms being developed in the fields of earth observation and robotics.