University of Bamberg
open search form
You are here

Information Systems and Applied Computer Sciences

Chair of Explainable Machine Learning

  1.  University of bamberg
  2. Faculties
  3. Information Systems and Applied Computer Sciences
  4. Subject groups
  5. Applied Computer Science
  6. Chair of Explainable Machine Learning
  7. In the Media
  8. AI-supported handwriting analysis: How a smart pen could improve learning right from the start
TV Oberfranken GmbH & Co. KG

AI-supported handwriting analysis: How a smart pen could improve learning right from the start

The university magazine 4you, published by regional television station TV Oberfranken, focuses on innovative projects for children and young people. One of the articles featured comes from the Chair of Explainable Machine Learning at the University of Bamberg. In a research collaboration with industry and education experts, the team is developing an AI-enabled pen designed to detect handwriting problems at an early stage. The goal: objective assessment, personalized feedback, and better learning opportunities for children.

Teaser

In its latest episode, 4you – the university magazine – presents innovative educational projects from universities in Upper Franconia. The focus ranges from creative theater formats and motor skills screenings to hands-on science activities for children. One contribution comes from the Chair of Explainable Machine Learning at the University of Bamberg, where researchers are combining artificial intelligence with a familiar everyday object to support learning at an early stage.

The report presents the KIEBEL research project – AI in education for the evaluation of legibility, which was launched at xAILab Bamberg. The project addresses a key challenge in education: difficulties in writing are often only recognized once they have already impacted school performance, while the assessment of legibility remains largely subjective.

In collaboration with the Schreibmotorik Institut in Heroldsberg and the writing instrument manufacturer STABILO, our team is developing a sensor-equipped pen that records movement data such as pressure, speed, and acceleration during writing. This data forms the basis for machine learning models that correlate measurable writing behavior with perceived legibility.

As Prof. Dr. Christian Ledig explains in the report, large collections of handwriting samples, evaluated by both experts and laypeople, are used to train the system. An important goal is to define transparent, objective criteria for evaluating handwriting so that the AI can provide individualized and understandable feedback.

In the long term, the smart pen will not only identify potential writing difficulties at an early stage, but also suggest tailored exercises that children can use to improve their handwriting in a playful way. This will directly benefit students, but also teachers in the classroom and parents, who will gain objective insights into their children's learning progress. It is important that the pen retains the look and feel of a conventional writing instrument so that the technology does not interfere with natural writing behavior.

In addition to the Bamberg project, the program also includes educational initiatives at Coburg University, the University of Bayreuth, and Hof University, which illustrate the breadth of research-based work with children and young people in the region. In this diverse environment, AI-supported handwriting analysis is an example of how explainable AI and interdisciplinary collaboration can make a concrete contribution to early childhood education.

See the full broadcast here

4you - das Hochschulmagazin: KI-Stift soll Schreibschwächen frühzeitig erkennen