The Austrian research project DOMINO kicked off in June as a countrywide initiative to advance the next generation of industrial AI applications. The goal is to create artificial intelligence directly on decentralized devices—known as EdgeAI—more secure, energy-efficient, and with increased permanent performance. The DOMINO flagship project addresses the current challenges that companies face in the practical use of AI in everyday life. These include the certification of AI in safety-critical areas, the recognition and evaluation of AI performance, and the involvement of users in AI decisions. Based on open-source code, AI solutions are developed on edge devices considering the complete DevOps lifecycle of Internet of Things (IoT) devices.
In addition, the project makes a significant contribution to strengthening the competitiveness of Austrian industrial companies in the field of IoT and beyond. using the DevOps method. The specific requirements of Austrian industry are directly incorporated into the design.
Why EdgeAI?
Edge computing, or data processing directly on the edge, e.g., on machines, has numerous advantages over centralized cloud solutions: it is faster, saves energy, offers better data protection, and can reduce costs. However, implementation is not easy: AI models must be adapted to the limited performance of edge devices, monitored during operation, and securely updated as needed. Running AI algorithms on resource-limited hardware requires in-depth expertise—particularly in the areas of model optimization, code generation, real-time processing, and system integration.
How artificial intelligence works reliably directly on the edge
The DOMINO project chooses a holistic approach by considering the entire life cycle of EdgeAI systems, from development to operation to updating. The three main objectives are:
- Development of AI-optimized models for industrial applications (railway, renewable energy, wood industry)
- Monitoring system performance under real-world conditions (e.g., dust, weather, equipment wear, etc.)
- Continuous improvement through operational data and user feedback (human-centered approach)
The aim is also to significantly reduce the development effort of Edge AI applications for companies, especially SMEs, while meeting the high requirements for safety and sustainability, as specified in the EU AI Act, for instance. The development of a standardized development cycle will also result in guidelines that will serve as a blueprint for companies that want to develop Edge AI solutions in the future.
The consortium
The project, led by TU Wien, brings together experts from industry and research. Siemens, Mission Embedded, and VRULL are involved as technology partners, while TU Graz is another research partner and FILL (wood processing) and Reisenbauer Solutions (photovoltaics) are industry partners. The project results will be implemented in three specific use cases and demonstrated in a practical setting. Platform Industry 4.0 Austria is responsible for the communication and dissemination of the project.
Involvement of external expertise
Through an “External Expert Board”, external stakeholders who are not directly involved in the project will have the possibility to interact with DOMINO. The board offers interested experts from business, science, and administration the opportunity to learn about current developments in the project and to actively provide feedback. A workshop halfway through the project will provide an opportunity for participation and knowledge exchange. If you are interested in participating, please contact: michael.faelbl@plattformindustrie40.at
The DOMINO project is funded by the Federal Ministry for Innovation, Mobility and Infrastructure (BMIMI) and the Austrian Research Promotion Agency (FFG).
