Spreading Machine Learning
The research results of the ML2R Competence Center are used by short routes for practice-oriented applications and make new data-based services possible. Application examples from the fields of industry 4.0 and logistics as well as the automation of cognitive processes provide orientation. ML2R offers, to small and medium-sized enterprises in particular, access to strategies, knowledge and data so that they can successfully apply Machine Learning technologies and hold their ground in international competition. The transfer offer ranges from concrete cooperation models such as the “Enterprise Innovation Campus” and a platform with curated data and models to the promotion of training and further education for specialists in Machine Learning.
Many companies have already recognized the potential of Machine Learning (ML) applications, but are facing key challenges: They need comprehensible, trustworthy technologies that they can flexibly integrate into their business processes. They lack specialists who develop the appropriate technologies and strategically implement them in the company. Often, there is a lack of well prepared data to optimally train learning systems and to use them profitably. In order to address these challenges, ML2R brings together excellent scientists with developers and users from businesses.
Scientists and Experts in the Think Tank - the Enterprise Innovation Campus
The partners of the competence center offer tried-and-tested models of technology transfer. The "Enterprise Innovation Campus" initiated by Fraunhofer IAIS has already proven itself as a think tank for application-related developments and continues to serve as a central forum for agile cooperation projects with companies in the future.
Industry 4.0, Logistics and Business Process Automation
Application areas of research at the ML2R are the topics Industry 4.0, Logistics and the automation and analysis of business processes. In concrete terms, it is for example a matter of improving production processes, drawing up forecasts for the quality of products or planning and controlling logistics systems. Data from networked machines and sensors can be used as practical training material for ML systems.
Science also benefits from this closeness to practical use: The ML methods and applications must be designed in such a way that they function in realistic scenarios and are able to cope with real, often heterogeneous data and complex knowledge.
Open Source Access to Data and Models
In line with the open source idea, the results of the research are made publicly available in the form of curated models, algorithms and data. In this context, curating means defining quality criteria for data and models, giving error estimates, describing application areas and documenting test procedures.
Finally, the ML2R Competence Center would like to contribute to the discussion on the future topics of Artificial Intelligence, Big Data and Machine Learning. To this end, it is planned to present the results of the research work in a scientifically sound and at the same time understandable form to the public.
In order to counteract the shortage of skilled professionals, the ML2R competence center is also involved in the development of an interdisciplinary training and further education program. For students, there will be courses and master's theses as well as opportunities for doctoral studies. Another important concern is to anchor knowledge on Big Data, Artificial Intelligence and ML methods not only in the field of computer science, but also to develop offers for engineers and legal and humanities scholars. Developers and users are also supplied with consulting and training modules, enabling them to design new data-based business models and successfully shape digital change in their companies. Finally, the ML2R Competence Center will protect jointly developed results in the form of patents and promote the creation of start-ups.