Work Packages

WP1: Project Management
WP1 will ensure an effective and efficient budget and time management, as well as the quality and content assurance, by implementing the most appropriate tools and means that will guarantee the monitoring of all activities, administrative management, coordination, budget implementation and communication between all the partners.

WP2: Elaborate Feedback and Help
WP2 enhances learning in MathCityMap through AI-driven, personalized feedback for students and advanced support tools for teachers. The work package focuses on capturing and analyzing students’ solution processes in order to generate meaningful, tailored feedback and actionable insights.
A dedicated communication interface in the web portal enables interaction with a generative AI chatbot. Trained on existing documentation and training materials, the chatbot provides on-demand guidance for teachers using the platform.
Within the app, a new upload function allows students to submit their solutions (e.g., photos or files). These submissions are processed by a coordinated system of specialized AI agents that identify errors, highlight missing steps, and suggest improvements. The resulting feedback is structured, transparent, and pedagogically meaningful.
For teachers, an adaptive feedback function aggregates common errors and provides privacy-conscious statistics to support instructional planning. The entire system is continuously evaluated with participating schools to validate its pedagogical effectiveness and enable targeted improvements.

WP3: Enriched Task Creation
WP3 streamlines and enriches the creation of mathematical modelling tasks by integrating AI-supported inspiration and guidance into the MathCityMap platform.
A new login function in the app securely connects user accounts between the app and the web portal, ensuring consistent ownership of uploaded images. Teachers can upload photos of real-world environments, where AI-based image recognition identifies relevant mathematical objects. Based on these detections, adaptive recommendation tools suggest suitable elements from an existing database to inspire task development.
At the core of this work package is the Universal Task Wizard, which replaces the previous wizard in the web portal. Using multiple relevant parameters, it generates structured task drafts based on measurements, images, and contextual information. This significantly lowers the threshold for creating high-quality modelling tasks without requiring extensive prior training.
An iterative evaluation process with partner schools ensures that user needs are systematically captured and that task creation remains intuitive, efficient, and pedagogically sound.

WP4: Sustainable System Modernization
WP4 modernizes the technical foundation of MathCityMap to ensure long-term sustainability, security, and scalability.
The legacy AngularJS frontend will be replaced by a newly developed Angular-based interface, including a complete UI redesign and necessary backend adaptations. The system will be fully GDPR-compliant and designed according to privacy-by-design principles, while remaining ready for seamless integration of AI-based features.
A new database structure and carefully planned data migration will ensure reliable and secure user management. To handle growing amounts of user-generated content, a semi-automatic AI-supported review process will pre-evaluate tasks so that only special cases require manual review.
The Digital Classroom will be significantly enhanced through advanced teacher views and privacy-first learning analytics. Aggregated, interpretable indicators presented in dashboards will support lesson planning and reflection—without profiling individual students.
Systematic evaluation with participating schools will ensure that the modernized platform effectively supports teaching, research, and future innovation.

WP5: Dissemination & Evaluation
WP5 combines dissemination and evaluation to ensure broad adoption, practical implementation, and long-term impact of the project results.
A central component is the development of an adaptive Massive Open Online Course (MOOC) and accompanying materials introducing the new AI-based features. These resources form the basis for extensive national teacher training activities in Germany, Austria, Estonia, Italy, Romania, Slovakia, and Spain. Through these trainings, teachers will be qualified to confidently use the new tools while contributing to the growth of an international community.
An international Teacher Camp and dissemination meeting will serve as a multiplier event, preparing sustainable dissemination strategies beyond the project’s lifetime. Teacher training materials will be continuously adapted and expanded to support effective knowledge transfer.
Finally, a coherent corporate design—including logo, templates, website, and Instagram channel—will ensure consistent visibility and communication across Europe.