Privately funded TUM research project develops SARAkids, an intelligent AI reading assistant
for children

Reading is fundamental to successful learning in all subjects and a central key to educational equality. Nevertheless, many children encounter obstacles at an early age that can slow down their learning progress. Different developmental trajectories can be just as impairing as limited support from the family environment. A research team at the Technical University of Munich (TUM) is tackling this challenge with SARAkids (Smart AI Reading Assistant for Kids): a scientifically based learning companion that recognizes reading difficulties in real time and provides targeted, motivating support.

Importantly, SARAkids is designed as an accompanying offer to support learning processes and to complement lessons, support, and family learning. The system does not replace teachers or proven pedagogical concepts and is not intended as a “reading aloud” tool or for “teaching automation”. Rather, SARAkids is designed to help children reconnect more quickly and develop strategies when practicing independently.

“Every child should have the chance to become a confident and self-assured reader,” says Prof. Dr. Enkelejda Kasneci (TUM Center for Educational Technologies), who is the scientific director of the project. “With SARAkids, we combine learning research, eye-tracking, and generative AI to provide support exactly when needed: individually, motivating, and scientifically carefully evaluated.”

The project has been made possible thanks to a private donation from Dr. Christoph Kuhmann, LL.M. “I support SARAkids because early reading success has a lasting impact on educational biographies,” he emphasizes. “If technology can support children precisely when they encounter difficulties, we can open up new opportunities for greater self-confidence, participation, and long-term prospects.” Dr. Kuhmann worked as a lawyer in Munich for over 30 years. His work focused on corporate transactions, most of which had an international dimension. He also lectured at the TUM School of Management for many years.

Personalized support – at exactly the right moment

“Thanks to Dr. Kuhmann’s support, we can tailor the development of SARAkids entirely to the needs of learners”, affirms Franziska Kaltenberger, who is doing her PhD as part of this project. After completing her studies, she worked as a teaching assistant at a secondary school. ” I have seen how children can benefit when they receive support tailored to their needs. This experience motivates me in my research.” She believes that technologies such as SARAkids not only offer opportunities to provide children with this individualized learning experience but also open up new ways of learning and teaching.

With SARAkids, for instance, learning support for beginning readers is no longer based solely on comprehension questions or reading aloud tests. Instead, the tool uses eye tracking: Based on real-time analysis of children’s eye movements, SARAkids recognizes when a child is experiencing difficulties with reading, for example, with recognizing a word or understanding a context. SARAkids then responds immediately with individualized interventions that can be visual, auditory, or multimodal, depending on the child’s needs. Support feedback is generated using generative AI, enabling an adaptive, personalized learning process during reading practice.

SARAkids illustrates how technology can promote educational equity. By providing personalized, real-time support, it ensures that every child has the tools to become a confident reader and learner, regardless of their starting point. Projects like this underscore the transformative potential of educational technologies in creating more inclusive learning opportunities for all.

More Information

https://www.edu.sot.tum.de/hctl/research/sara

Contact

Franziska Kaltenberger (franziska.kaltenberger@tum.de)

Prof. Enkelejda Kasneci (enkelejda.kasneci@tum.de)