Dikes protect against floods and are thus part of Critical Infrastructure (CRITIS) in Germany. A failure of dike structures is associated with dramatic consequences such as fatalities, economic and social damage, but also damage to ecologically relevant natural areas, and is becoming increasingly relevant because of climate change. Many system-relevant dikes do not have any integrated sensor technology for monitoring, but in practice, only the hull can be inspected during regular dike inspections. This is still a process today without knowledge of the processes inside the dike: Human dike runners observe the dike in regular routine inspections as well as in the event of stress and examine it for damage (e.g. seepage or erosion) through primarily visual inspections. In the research project “Digital Dike Runner”, the traditional dike runner is being replaced by innovative technologies and digital methods in order to (semi-)automatically, efficiently, and ecologically minimally invasively record dike systems and faulty and damaged areas using “X-ray vision”. The aim is to record the current internal structure, including damage and faults, using ground penetrating radar (GPR) technology mounted on drones. For (semi-)automatic evaluation and assessment, methods of artificial intelligence (AI), such as machine learning, are to be adapted and applied to the radargrams. The lecture gives an overview of the almost completed investigations and developments.