Projects
Ongoing projects
BlitzKorrektur - 3 Decades of Lightning Data under Scrutiny: Machine Learning corrects Distortions through Measurement Network Evolution
Lightning strikes cause significant damage to people, infrastructure, and buildings, trigger wildfires, and force cancellation of outdoor events. Lightning is the only meteorological variable measurable continuously in space and time across large areas – achieved for three decades through Lightning Location Systems (LLS). However, these datasets have a critical limitation: measurement systems have been continuously upgraded. New sensors, modified configurations, and improved algorithms influenced measurements but quantitative impact remains unknown. Initial investigations suggest significant artifacts compromising reliable use of these valuable long-term datasets. This university-industry collaboration will quantitatively assess and correct measurement network evolution impacts. Using meteorological data and machine learning, hidden artifacts will be identified, modeled, and removed from time series.
Principal investigators:Reto Stauffer, Georg Mayr
Funding agency: FFG
Project partner: OVE Service GmbH
Duration: 2026-2029
GROUPY - Zero-shot multitask models for analyzing group-based rhetoric in political texts from synthetic labeled data
This project develops GROUPY, a suite of reliable and scalable automated methods for identifying and categorizing social group mentions in political texts using advanced natural language processing (NLP) methods. Political competition, representation, and polarization are deeply rooted in divisions along group lines. Studying when, how, and why politicians, parties and other political elites talk about social groups can offer intriguing insights into the connection between electoral campainging, political representation, democratic governance, and societal solidarity and polarization. Yet, despite the importance of group-based rhetoric for understanding political competition, representation, polarization, and public opinion formation, existing computational methods for analyzing it lack scalability, reliability, and/or generalization.
Principal investigator: Hauke Licht
Funding agency: TNF
Duration: 2026-2027
BRIDGE - Building Research Insights with Data and Graph Exploration
The primary goal of this project is to develop an interdisciplinary research database that integrates diverse datasets and facilitates advanced analyses. This platform will act as a dynamic tool, enabling researchers to quickly explore new research topics and test ideas with minimal effort. By consolidating heterogeneous data into a unified system, the project aims to foster interdisciplinary collaboration and drive scientific innovation. One of the core objectives is to encourage collaboration among researchers from different scientific disciplines. By providing a shared platform, this database will promote an integrated approach to research, breaking down traditional disciplinary barriers. Researchers will be able to explore connections between datasets, triggering new hypotheses and insights that were previously inaccessible due to data silos. The platform is designed to go beyond simply storing data. It will act as an incubator for new research questions and data-driven discoveries. Using advanced artificial intelligence, particularly large-language models (LLMs), the system will enable automated metadata generation, dynamic queries, and intelligent analysis of diverse datasets. This capability will significantly reduce the effort required to formulate and test ideas, opening the door to innovative methodologies and findings.
Principal investigator: Gerald Hiebel
Funding agency: Land Tirol
Project partner: Institute for Interdisciplinary Mountain Research (IGF, ÖAW)
Duration: 2025-2027
ELSA - Effective Learning of Social Affordances for human-robot interaction
Affordances are action opportunities directly perceived by an agent to interact with its environment. The concept is gaining interest in robotics, where it offers a rich description of the objects and the environment, focusing on the potential interactions rather than the sole physical properties. In this project, we extend this notion to social affordances. The goal is for robots to autonomously learn not only the physical effects of interactive actions with humans, but also the humansʼ reactions they produce (emotion, speech, movement). For instance, pointing and gazing in the same direction make humans orient towards the pointed direction, while pointing and looking at the finger make humans look at the finger. Besides, scratching the robotʼs chin makes some but not all humans smile. The project will investigate how learning human-general and human-specific social affordances can enrich a robotʼs action repertoire for human-aware task planning and efficient human-robot interaction. It is carried out in cooperation with ISIR, Sorbonne University and LAAS-CNRS, France.
Principal investigators: Justus Piater, Matthias Schurz, Erwan Renaudo
Funding agencies: Austrian Science Fund (FWF) and the French National Research Agency (ANR)
Project partners: ISIR, Sorbonne University; LAAS-CNRS, France
Duration: 2022-2026
Completed projects
ArtiPro - Artificial intelligence for personalized medicine in depression
The aim of this project is to establish an artificial intelligence platform that captures, integrates, analyzes and harmonizes data from clinical research on biomarker signatures and therapeutic outcome from high-quality multidisciplinary sources with the purpose of identifying robust multimodal biomarkers and outcomes for affective disorders. The platform shall focus on the area of antidepressant treatment in affective disease and the definition of functional imaging biomarkers in combination with molecular data (transcriptomics and genetics).
Principal investigators: Roberto Viviani, Clara Rauchegger
Funding agencies: ERA PerMed (ERA-Net Personalized Medicine), European Commission and FWF
Projekt partners: Diakonhjemmet Sykehus - Diakonhjemmet Hospital; IRCCS Istituto delle Scienze Neurologiche di Bologna; Zagrebu - University of Zagreb (UniZG); Tel Aviv University (TAU); Universitätsklinikum Aachen, AöR
Duration: 2022-2026
DIGIdat - Digital data analysis of indoor air quality meets ESD
Indoor air quality in schools has become a highly discussed topic in the wake of the Corona pandemic. But what is the best way to ventilate a classroom? To what extent does installing a ventilation system improve indoor air quality, and how do comparatively simple interventions (CO2 traffic light signal, awareness raising, etc.) fare? In the research project DIGIdat, researchers from the University of Innsbruck and the University College of Teacher Education Tyrol are working with students and their teachers to investigate the interplay between indoor air quality, thermal comfort, energy efficiency and awareness raising. Employing a Citizen Science approach, the project intends to actively involve a total of about 750 students at 10 Tyrolean schools in collecting measurement data.
Principal investigator: Gabriel Rojas-Kopeinig
Funding agency: OEAD - Sparkling Science
Project partners: University College of Teacher Education Tyrol (PH Tirol); BINK Initiative Baukulturvermittlung für junge Menschen; komfortlüftung.at; openSenseLab gGmbH
Duration: 2022-2025