Enabling chronically ill schoolchildren to take part in lessons thanks to mobile robot — this is the aim of a new project by the universities of Bonn, Göttingen (coordinator) and Duisburg-Essen in collaboration with the chilli mind GmbH. The particular challenge here is to protect the privacy of all those involved.
Although children and young people often moan about school: Everyday school life brings them into daily contact with their peers, which is not only important for their well-being, but also for their social development. Children with chronic illnesses currently often miss out on this. This is where the “Privacy-friendly mobile avatar for sick schoolchildren" (PRIVATAR) project, in which the Humanoid Robots Lab at the Institute for Computer Science Bonn is involved, aims to provide a remedy. The idea: mobile robots should enable sick children to take part in lessons and everyday school life. The robots are used as avatars at school, acting as legs, eyes, ears and mouthpieces for the sick children at home. The big challenge here is protecting the privacy of classmates, teachers and the sick children themselves.
"Robots are equipped with a wide variety of sensors so that they can move safely in their environment and interact with people,” explains Prof. Dr. Maren Bennewitz, Head of the Humanoid Robots Lab. Sensors detect furniture, for instance, but also movements or faces so that the robot can react to them. The recorded data also includes a lot of data relating to people's privacy, such as video recordings, location, or movement patterns.
"Our goal is to develop mobile robots that use as little sensor data as possible to ensure the privacy of the people in the classroom, but also offer safety and efficiency in handling," says Maren Bennewitz, who is also a board member of the PhenoRob Cluster of Excellence and a member of the Transdisciplinary Research Area "Modeling" at the University of Bonn.
In her sub-project, Bennewitz and her team are using their expertise to make robot navigation user-friendly for children and teachers — in particular how sensors can be switched on and off according to the user's preferences. "For example, we are planning to design the controls in such a way that children and teachers can make the desired settings without instructions. For example, they can change the privacy settings or specify how close the robot can get to them,” says doctoral candidate Nils Dengler, who is working on the project.
The particular challenge here is to ensure the functionality of the mobile avatar, even if individual sensors are temporarily or permanently switched off or their temporal or spatial resolution is reduced. "In these cases, it must be ensured that users are made aware that the robot's function may be limited."
The interfaces developed in the project will ultimately be integrated into an overall system for controlling the mobile robot in cooperation with the partners. The "data-saving" robots offer potential applications beyond the school context, for example in museums.