14 new projects are being funded by ZHAW digital as part of the Digital Futures Fund (DFF). These include projects for internal digital transformation and innovative research projects.
The "Digital Futures Fund" funding program from ZHAW digital is aimed at all ZHAW employees who want to develop or test an innovative idea for digital transformation in the fields of education and research as well as management and support areas. All project leaders will be automatically included in the «Digital Futures Lab» digital community through their sponsored project, allowing them to connect and exchange ideas more intensively with colleagues on specific topics. More than 50 projects have already been funded since 2020.
The most recently funded projects were divided into "Innovation Projects" and "Impact Projects".
Can Generative AI Simplify Complex Discrete Process Modelling?
Lukas Hollenstein. Advancing the digitization of complex process modelling in life sciences, our project explores AI's potential in generating process diagrams and models. This research aims to simplify modelling procedures, enhance simulation accuracy, and facilitate the development of efficient digital twins.
Learning Copilot - A Learning Assistant based on Large Language Models
Jochen Wulf. We develop a Learning Copilot that utilizes large language models (LLMs) like ChatGPT to assist students in understanding lectures and preparing for exams. The Copilot leverages a knowledge base of lecture transcriptions, papers, and books. We aim to evaluate its effectiveness in two ZHAW courses.
Multimodal Anonymization of Gameplay Data
Elena Gavagnin. Gameplay data is used to study human behaviour across disciplines. Unfortunately gameplay corpora are not anonymized, limiting their use for open research. In this project we use transformer-based multimodal AI methods to anonymize visual and textual gameplay data, while preserving game dynamics.
Personnel Guidance for Safeguarding Events through Faster Than Real-Time (FTRT) Simulations
David Bernhardsgrütter. The aim is to further develop software for simulating pedestrian flows, allowing calculations to be performed faster than real-time. The tool enables early detection of capacity constraints at events and can be used for dynamic control of visitor flows.
SmartLabHub - Remote Laboratory Data Infrastructure
Oliver Döbrich. The project aims to establish a digital remote infrastructure for lab environments at ZHAWs IMPE, leveraging the Internet-ofThings to connect lab equipment with advanced algorithms. The proposed system ensures low-threshold access for scientists, providing a foundation for Industry 4.0 initiatives.
TinyML Grasshopper Classifier: Enabling Non-Invasive Biodiversity Monitoring
Tobias Peter. Using AI to classify insect sounds, particularly grasshoppers, is a promising method to monitor biodiversity non-invasively in the field. We propose a sustainable tiny machine learning model to efficiently classify grasshoppers in real-time using edge computing with minimal computational resources.
Towards Enhancing Large Language Models with SNOMED CT for Multi-document Patient Records Summarization: A feasibility study
Ahamad Aghaebrahimian. Clinical physicians spend about 40% of their work time for reading and writing patient documentation. We will employ NLP, SNOMED CT, and Large Language Models (LLM) to generate concise, accurate, and interoperable summaries of patients’ records, thus saving time, effort, and resources.
Transforming Clinical Assessments: Explicitly Articulating Implicit Clinical Decision-Making to Train AI
Lena Sauerzopf. Researchers are involving artificial intelligence in assessments, using therapist ratings to create a ground truth. The project aims to assess the reliability of video-based observations for compensatory movements post-stroke to promote Artificial Intelligence (AI) assisted rehabilitation.
Application of ChatGPT and Co. in Teaching and Research at ZHAW
Christian Rapp. In a pilot study conducted in the summer of 2023, we investigated the use of AI in the creation of bachelor's theses across four departments at ZHAW. In this project, we aim to refine the study, conduct it across as many departments as possible, and expand it to include the perspectives of instructors.
Fair DFF Voting Design
Florian Spychiger. The voting process for DFF-proposals has been a point of much discussion. We aim to improve the voting mechanism to account for partiality based on relationships, campaigns and mobilization activities by compiling the relevant theory of digital democracy and suggesting a new mechanism.
Hack4SocialGood 2024 – Promoting Digital Inclusion
David Lätsch. Hack4SocialGood, a 2-day hackathon, bridges tech and social sectors, fostering collaboration to tackle digital challenges in social work. It unites diverse experts to create digital solutions for real-world social challenges, enhancing organizations` impact and promoting digital literacy.
Net Zero: A Computer Simulation Game for Sustainable Cities
Andri Gerber. Imagine a game where you are responsible for an average swiss city in 1990 and to win you have to reach the zero emission goals by 2050. This is your role in a game that handles the influence of the construction industry on co2 emission and waste production. The game is developed for schools.
Immediate Help for School Stress: Digital Feedback System for Adolescents
Anthony Klein Swormink. 40% of Swiss adolescents experience some degree of stress from school. We are testing with them to what extent wearables, in combination with an app, help to better detect and cope with stressors and states of stress. Stress is measured in real-time, and they receive immediate feedback and tips.
ZHAW Research Tinder - Find Your Research Partners
Kevin Andermatt. Through ZHAW Research Tinder (working title), we provide full access to the combined brainpower of the university. With this browser app, researchers can easily and playfully connect across organizational boundaries, finding the right partners for research projects.