DIG-IT!: Digitization of natural complexity to solve ecological problems that are relevant for society
Prof. Martin Wilmking, Ph.D.
Dr. Mario Trouillier
Phone.: +49 (0)3834 420 4185
The excellency program of the federal state of Mecklenburg-Vorpommern for the digitization of research is funding the project ‘DIG-IT!' Digitization of natural complexity to solve ecological problems that are relevant for society’, lead by Prof. Martin Wilmking. From Juli 2019 to August 2022, two Mio. Euro are available for the researchers to develop a methodological toolkit to analyze ecological image and sound data with machine learning methods (‘deep convolutional neural networks’). Several institutes from the University Greifswald are part of DIG-IT!, as well as Fraunhofer Institute for Computer Graphics Research IGD in Rostock.
The aim of DIG-IT! is to grasp the opportunities digitization offers and make the new technologies applicable for ecological research. Thus, urgent ecological questions that are relevant for society will be tackled with state-of-the art technologies, while also qualifying a new generation of digitally competent ecologists, bio-mathematicians and computer scientists. A wide range of research topics is explored, for example ecosystem stability under climate change and land-use change, species protection, as well as innovative environmental monitoring. In particular, the aim is to facilitate a quantum leap in various disciplines by developing universally applicable methods that use trainable algorithms (deep convolutional neural networks), because in the digital age the main challenge is not just the amount of data, but also the evaluation of vast databases. To this aim, DIG-IT! connects the expertise of the developers form the Fraunhofer Institute for Computer Graphics Research and the Institute for Biomathematics at the University Greifswald with the applied research questions regarding urgent ecological questions which are explored in the Institute for Botany and Landscape Ecology and the Zoological Institute and Museum.
Publication statistics for ‘deep learning’ between 2002 and 2017 (research by the Fraunhofer IGD by using Scopus). Biology (‚agricultural and biological sciences‘ in ‚other‘, right) only contributes about 1% of all these documents – thus DIG-IT!
Extract from the jury‘s vote:
„The proposal DIG-IT! aims to tackle urgent ecological questions by the application of digital tools, in particular the evaluation of primary data, which currently poses a challenge. (…) The jury is convinced of the projects scientific excellency. In particular the idea to develop a digital toolkit is impressive. Appreciated is also the connection to previous and existing research cooperatives like WETSCAPES und RESPONSE [DFG graduate college ‚Biological RESPONSEs to Novel and Changing Environments‘] and to use their findings for the promotion of the young academics.“
The project "DIG-IT!" is funded by the European Social Fund and the Ministry for Education, Science and Culture of the federal state Mecklenburg-Vorpommern.