The establishment of the Centre for Computational and Dat Science (dScience) was approved by the Faculty Board at the Faculty of Mathematics and Natural Sciences (MN) on June 22, 2020. dScience was put into operation on January 1, 2021 and officially opened by Minister Bj?rn Arild Gram, on October 27, 2021.
The first year has been affected by the pandemic, but has largely been carried out as planned. Our graphic profile and websites have been established, as well as the premises in the Penal in Kristine Bonnevie's house (entrance through Georg Morgenstierne's house). A physical meeting place has been built, which will mainly be used by PhD and postdoctoral candidates across UiO. A lot of work has been done in connection with the development of the premises in 2021.
Role and tasks
Throughout 2021, management, professional councils and the board have worked out an overall description of the role that dScience will play, at UiO, nationally and internationally. The role of dScience can briefly be described in the following three points;
- dScience will further develop the University of Oslo as an internationally leading research-intensive university in data and computational science.
- dScience must be a driving force for, including prioritization, activities that contribute to green transformation and a sustainable future – for all.
- dScience will develop and maintain mechanisms for collaboration between academia, business and the public sector, including being a driving force in the development of Oslo Science City.
The assignments, given by the faculty and approved by the board, are of course related to the role described above and can be summarized as;
- dScience will develop and run meeting places and events across departments and faculties at UiO and in collaboration with partners in Norwegian society and business, including developing and further developing national and international network collaboration.
- dScience will, together with USIT, UB and other service providers, develop the infrastructure necessary to carry out high-quality research and education in data and computational science.
- dScience will facilitate project collaboration and develop projects across disciplines and units at UiO and in collaboration with partners in society and business.
- dScience will develop long-term strategic collaboration with selected partners in business and the public sector.
- dScience shall contribute to education in general and continuing and further education in particular.
Projects and project work
When starting up, dScience was assigned the responsibility for allocating 12 KD-positions (PhD) - 6 in 2021, 2 in 2022 and 4 in 2023. The first announcement resulted in dScience allocating 6 for 2021 and that the Department of Informatics used the opportunity to allocate 2 extra based on the 78 proposals that came in from the communities. These 8 projects initiated in 2021 are (host institute in parentheses):
- Bayesian Machine Learning for Complex Systems: From point estimates to uncertainty predictions in nuclear astrophysics experiments (Department of Physics)
- Distributed Machine Learning for Data and Computation-intensive Tasks (Department of Informatics)
- Improving predictions in histopathology using semi-supervised learning under the presence of domain shifts (Department of Informatics)
- Inductive bias for more efficient language modelling (Department of Informatics)
- On the Barriers of AI in Scientific Computing (Department of Mathematics)
- Online Anomaly detection in high-dimensional data streams – contextual anomalies and sensor networks (Department of Mathematics)
- Structure-Aware Machine Learning to Digital Twins (Department of Informatics)
- Swarm intelligence for observing systems in climate science – developing a reinforcement learning framework for surface flux mapping with drones (Department of Geosciences)
Furthermore, in consultation with MN, it has been decided that one ?dScience? position from the 2022 quota should be allocated to each of these projects (host institute in parentheses):
- PriTEM: Privacy preserving Transactive Energy Management (Department of Informatics)
- EMPOWER: Sustainable Batteries in Mobility - (Em) powering a Net-Zero Energy Transition (Department of Technology Systems)
During 2021, it also became clear that UiO had two applications in the final of the SFF scheme, both of which are in their entirety within dScience's professional field. If these are approved by the Research Council of Norway, they will receive two dScience KD positions each in 2023. These applications are (host institute in parentheses):
- Twins4Life - The Science of Digital Twins (Department of Informatics)
- Integreat - Norwegian centre for knowledge-driven machine learning (Department of Mathematics)
Furthermore, the following two SFF applications have significant elements of data and computational science (host institute in parentheses):
- ACT - Centre for Arctic Climate Transition (Department of Geosciences)
- BrightMatter: the Quest for the Origin and Evolution of Heavy Elements (Department of Physics)
One position has been allocated to each of these from 2023, if they are granted by the Research Council of Norway.
Seminars and events
An important part of dScience, is to create meeting places, both physically and digitally. This part has naturally been affected by the pandemic, but dScience has nevertheless delivered well. The following is an overview of our own events and a little about what dScience has participated in.
Lunch seminars for PhD candidates and postdocs at dScience's premises
- October 21: Presentation of SIRIUS, David Cameron (IFI)
- October 28: Presentation of NORA, Klas Pettersen (NORA)
- November 4: Presentation of BigInsight, Ingrid Glad (MI)
- November 11: "Uncertain Energy Systems", Fred Espen Benth (MI) and Marianne Zeyringer (ITS)
- November 18: ?In high-dimensional parameter space no one can hear you scream: Searching for new physics with smart sampling and machine learning?, Anders Kvellestad (FI)
- December 16: Discussion: upcoming program 2022, Ingrid Glad and Morten D?hlen (on Zoom)
Lunch seminars in the Science Library, open to everyone
- October 14: Kickoff, first event in dScience's lunch seminar series. "Deep Learning and Neural Networks: an infinite dimensional perspective", Luca Galimberti (NTNU) and introduction of dScience, Morten D?hlen (dScience)
- November 25: "The Science of Digital Twins", David Cameron (IFI)
Digital Resources: dScience webinar series
- June 2: ?Digital Resources at dScience?, Morten D?hlen and Gard Thomassen (USIT), first event in dScience's webinar series on Zoom
- September 22: ?Available Digital Resources Nationally and Internationally?, Gard Thomassen, Kristen Nygaard's house (streamed on Zoom)
- December 8: "Data Protection in Research", Vilde S?rb? Nenseth (USIT)
Seminars and annual events
- September 9-10: dScience Strategy seminar, Voksen?sen Hotel
- October 27: Official opening of dScience at Data Science Day @ UiO 2021 in the Science Library and Sophus Lie's auditorium with academic presentations ?Why uncertainty matters in the age of deep learning and big data?, Simen Eide (Finn.no) and ?Standing on the shoulders of data: Current research in Language Technology?, Lilja ?vrelid (IFI)
- December 2-3: ?PhD Program Seminar 2021?, seminar for PhD candidates affiliated with dScience and CompSci, Olavsgaard Hotel
Other
- February 15: ?Digital twins and sustainability: Double gain with reuse of data?, webinar under the auspices of DigitalNorway (co-organizer, digital)
- April 29: ?Digital Life Norway meets dScience - how can Norwegian biotechnology benefit from cross-centre collaboration??, webinar under the auspices of Digital Life Norge (co-organizer, digital)
- August 31-September 1: AlphaFold v2.0 and RoseTTAFold workshop, webinar under the auspices of NORA (co-organizer, digital). Both AlphaFold and RoseTTAFold were installed on UiO's machines and were used during and after this seminar.
In addition to our own events, dScience has participated and been presented at a number of seminars and webinars during 2021. Information about the development of dScience and posts about the activity have also been published in writing in various channels, mainly on UiO's website, in Titan.uio.no and Khrono.
Digital resources
An important task for dScience is to develop and facilitate efficient utilization of digital resources, i.e., data, software systems and expertise with associated cost-effective capacities for data storage and access to necessary computing power locally, nationally and internationally. dScience has entered into an agreement with USIT, and Gard Thomassen, who heads IT in research at USIT, is engaged 20% to lead digital resources in dScience.
In 2021, the work has mainly been about planning, application writing and information work. Concrete projects, with funding, have been prepared and will start in 2022. One of these is support for the establishment of the national infrastructure Norwegian Artificial Intelligence Cloud (NAIC). The Research Council of Norway has confirmed that the application to establish NAIC has been approved. UiO will host this infrastructure and dScience will have a role in the implementation of the project.
There is a lot that can be mentioned under this area, but the most important thing that has happened is probably the establishment of the collaboration platform Educloud Research, the establishment of the FOX computing system, as well as necessary extensions and improvements of Services for Sensitive Data (TSD). This work has been carried out by USIT in close dialogue with dScience. dScience has also been involved in the FAIR@UiO project (FAIR-Findable, Accessible, Interoperable, Reusable), in particular how this work can be linked to collaboration with external partners.
There is great interest in the work with digital resources also among partners in business and the public sector.
Other initiatives and tasks
Through 2021, dScience has, by center manager, been involved in the development of Oslo Science City. The visions around the meeting places of the future ("co-working spaces"), prepared in dScience over the past year, were included in the feasibility study presented by Oslo Science City on November 30, 2021, see photo showing a sketch of a new building at Blindernveien station. This work will continue in 2022.
Through 2021, dScience has, by center manager, been involved in the development of the Norwegian Artificial Intelligence Consortium (NORA). This work will continue in 2022.
dScience is, by center manager, responsible for one of the courses in the new micro-education program ?From data to insight?. This course, entitled "The value of data in decision-making processes", was first completed in December 2021. The course was fully booked with participants from Norwegian business and the public sector. The course will be given again in the autumn of 2022.
dScience's main task is to develop basic long-term research in data and computational science across disciplines and sectors. However, there is great interest in these areas in the business community and in units in the public sector. The interaction with clusters, companies and units in the public sector has therefore been extensive in 2021. This work will continue in 2022.
Staff, Board, Council and Committee
Center Management
- Morten D?hlen (started 1.1.2021)
- Ingrid Glad (started 1.10.2021, acting center manager, part-time)
Community Services and Daily Operations
- Eva Michelsen Ekroll (participated in the establishment of dScience from the beginning, but formally started in dScience 1.5.2021, parental leave from 1.9.2021)
- Cecilie Linea Ellefsen (started 30.8.2021)
Digital resources
- Gard Thomassen (started 1.5.2021, 20% position)
Project / Program Development
- David Cameron (1.8.2021, funded by the SFI SIRIUS at the Department of Informatics, Prepare investment in digital twin technology and contribution to the development of strategic partnership program, part-time)
- Arne Bang Huseby (started 1.8.2021, funded by the Department of Mathematics, Head of work with COFUND application, part-time)
Support Services
The entire faculty administration and an attentive administration at the Department of Biosciences have contributed to the development of dScience in 2021. Jarle Nygard (development of premises), Hilde Hvistendal (research administration), Therese Ringvold and Torunn Guttormsen (HR), Ivana Vujic-Pavicevic (economics), Ida Marie Bj?rknes (communication) and Linn Katrine Hirsti (Science Librart) have contributed particularly much throughout 2021.
The Board
- Solveig Kristensen, Chairman of the Board (Faculty of Mathematics and Natural Sciences)
- Morten Dalsmo (SINTEF Digital)
- Astrid Fossum Gulbransen (DNB)
- Stephan Oephen (Department of Informatics)
- Geir Dahl (Department of Mathematics)
- Susanne Viefers (Department of Physics)
- Cecilie Rolstad Denby (Department of Technology Systems)
dScience Council
- Anne H Schistad Solberg (Department of Informatics)
- Einar Broch Johnsen (Department of Informatics)
- Geir Kjetil Sandve (Department of Informatics)
- Erik Velldal (Department of Informatics)
- Ingrid Glad (Department of Mathematics)
- Arne Huseby (Department of Mathematics)
- Sergiy Neshveyev (Department of Mathematics)
- Arnoldo Frigessi (Faculty of Medicine)
- Are Raklev (Department of Physics)
- Trude Storelvmo (Department of Geosciences)
dScience Community Forum
- Hans Arnold Winther (Department of Theoretical Astrophysics)
- Ann-Cecilie Larsen (Department of Physics)
- Simen Kvaal (Department of Chemistry)
- Norbert Pirk (Department of Geosciences)
- Jonas Paulsen (Department of Biosciences)
- Marianne Zeyringer (Department of Technology Systems)
- Ida Robertsen (Department of Pharmacy)
- Alexander Binder (Department of Informatics)
- Egor Kostylev (Department of Informatics)
- Kjetil Lysne Voje (Natural History Museum)
- Louise Emilsson (Faculty of Medicine)
- Diana Saplacan (Department of Informatics)
- Salvador Ortiz-Latorre (Department of Mathematics)
- Mikael Mortensen (Department of Mathematics)
- Johan Pensar (Department of Mathematics)