Medical Assistance Systems

At the IEEE International Conference on Development and Learning (ICDL) 2025, he presented a model that enables the robot Nao to flexibly adapt its explanation strategies to the users behavior and presumed cognitive states. Using gaze data, Nao decides when negations—such as “don’t tilt, but shake”—are helpful. In a study, this adaptive communication reduced the error rate by 23%.

The code is available as open source.

Contributors: André Groß, Dr. Birte Richter, Prof. Dr. Britta Wrede Application in the context of medical tasks on a touchscreen