Project

Infographics are powerful tools in industrial contexts, effectively conveying complex information in an appealing and easily comprehensible manner. For instance, they can be used to visualize production workflows, metrics like OEE, rules, hazard warnings, or storage layouts. They allow information to be absorbed and understood more quickly, complex data to be structured and presented clearly, and visual presentations to capture attention and interest effectively.

Furthermore, graphical formats intuitively represent data, enhancing communication, engagement, and collaboration by making data accessible and actionable. However, visual presentation methods vary with purpose, scope, and audience. While InfoVis typically produces interactive visual representations for professionals with a scientific focus, infographics are more engaging and visually striking.

However, the creation of tailored infographics is still time-consuming, and oftentimes they are only used once, which is inefficient. Here, GENIUS comes into play: We propose a novel approach that employs generative AI and Large Language Models (LLMs) to analyze in-house text corpora in PDF format and generate appropriate infographics from reliable sources. Our primary target user group for the software comprises project developers and consulting engineers, who frequently undertake consulting activities and rely on infographics to clearly present information during client discussions.

The primary goal of the GENIUS project is to accelerate the creation of easy-to-understand infographics while improving output quality. This is realized through an LLM-powered software prototype designed to assist users in the automatic creation of infographics. Compared to state-of-the-art methods for creating infographics, the proposed solution enables the analysis of large document collections, provides reliable insights and relationships between entities, and actively involves users in the graphical design process.

Fraunhofer Austria Research GmbH – Center for Data Driven Design (FhA DDD) and the Fraunhofer-Gesellschaft e.V. – Fraunhofer Institute for Systems and Innovation Research (ISI) are the participating partners in this project.

July 1, 2025 to June 30, 2026

• Fraunhofer Austria Research GmbH, Vienna, Center for Data Driven Design, Graz