In-house innovation

ODF Sweden is Sweden’s national ocean data lab, and ODF’s mission is to enable data-driven innovation by both commercial and non-commercial actors to ensure that the ocean and its resources are managed in the best possible and most sustainable way.

ODF Sweden focuses on solving global and local challenges through the application of technologies such as Machine Learning (ML) to different types of ocean data. Very often we work with data from international data aggregators such as EMODnet and Copernicus Marine Service

Fast Innovation Cycles & Data Bakery

ODF Sweden’s primary activity is to conduct ocean data-driven innovation challenges of approximately six months duration. We refer to these as our innovation cycles. Each challenge is focused on a use case and involves internal consortium work, open workshops and events, and the open dissemination of results and knowledge gained during the challenge. We illustrate this process by the ‘innovation wheel’.

However, although this process delivers tangible results within six months, it also kick-starts projects with longer time frames, which ODF Sweden increasingly focuses on. We have used the ‘Data Bakery‘ concept as a method to come up with new ideas, and to make these ideas as concrete as possible. See for example this idea about finding ways to deal with security sensitive data, from our Fifth Cycle – The Data Availability Challenge.

Seventh cycle - The do-it-ourselves Challenge

In our seventh cycle (starting in January 2023) we step a bit outside of our comfort zone… ODF Sweden has typically focused on data analysis, but this time we have decided to rattle several links of the data-knowledge-action chain. The core activities in this cycle are to build and deploy ocean sensors. A lead partner in this endeavor is Chalmers with Chalmers Revere Lab and Chalmers Fuse. Chalmers will also mobilise 110 second year students for the challenge ‘Filling the coastal data gaps!’. We make use of existing efforts into low-cost sensors, such as the OpenCTD project.

An important purpose of this cycle is to make further progress in previous cycles. In cycle #5 (‘the data availability challenge‘)  we learnt the need to take a loooong perspective on data availability and FAIRness (Findable, Accessible, Interoperable, Reusable). Our partners SND and SMHI are experts here. In cycle #6 (‘the citizen science challenge‘) we learnt about reaching out and making partners with new groups and individuals, to leverage our own and other’s competence and resources. Adlerbert professor in marine citizen science Uta Wehn provided crucial expertise here. 

Sixth cycle - The Citizen Science Challenge

In our sixth cycle (starting in July 2022) we take on the challenge of opening up the scientific process itself, going beyond academic institutions and company R&D departments to involve interest groups and interested individuals. Broadening the knowledge creation process – often referred to as ‘citizen science‘ (closely linked to ‘open science‘) – is rapidly gaining importance all over the world. ODF Sweden believes strongly in this development and we are determined to be a facilitator for marine citizen science. Citizen science is not new to ODF Sweden; all our previous innovations cycles  have actually been open to involvement. However, now we want to focus more specifically on both the benefits and rewards that can be won from the broad co-creation of knowledge, in terms of better data sampling, new scope, and better on-the-ground impact, but also the potential challenges around issues of privacy, trust, and streamlining data flows.

Fifth cycle - The Data Availability Challenge

This cycle represent the start of our second mandate period from the Swedish Innovation Agency (Vinnova): To continue as the national data lab for the marine/maritime domain.

In our fifth cycle (starting in January 2022) we take on a challenge which has been us from the start: Data availability. What ocean data is actually out there, and what is not? How can available data be accessed? Formats? APIs? Real time or historical data? These question are fundamental, and often the answer (and the devil) is in the detail. As any AI specialist will tell, finding and preparing the data is at least 80 % of the work when developing a model.

Fourth cycle - The Growth Challenge

In our fourth cycle (starting in January 2021) we approach the end of our initial mandate period from the Swedish Innovation Agency (Vinnova): To be the national data lab for the marine/maritime domain. During this cycle we have decided to focus on two things:

  1. Inviting all to discussions about digital innovations in the Blue Economy. Please join ODF Sweden’s Webinar Series! Sometimes the topic is technical and sometimes it is… something else. Discussions are always inspiring, because ODF Sweden is a motley crew with a shared passion for the ocean.
  2. Gearing up for the next level of support for the digital Blue Economy. Our initiative Ocean Digital Innovation Hub has been shortlisted by the Swedish government as a candidate to become a European DIH, with full innovation support to companies and the public sector.

Third cycle - The Killer Algae Challenge

For its third innovation cycle (starting in July 2020), ODF Sweden has developed the following challenge: To develop an AI-based prognosis tool to predict where, when and how Harmful algal blooms (HABs) will form in the context of aquaculture and tourism. The AI tool should predict HABs both short-term (days to weeks) and long-term (years to decades), given the climate change projections for Swedish waters. We are working with two case studies to develop HABs prognosis models on a subset of data and specific species of microalgae in: (1) the aquaculture industry on the Swedish west coast (the Skagerrak Sea) and (2) the tourism sector in the Baltic Sea. Extensive ocean data for these essential variables, such as those above, can also be found in databases open to the public.

By targeting a more general project aim and by looking into specific case studies, ODF Sweden is answering the needs from several different end users and stakeholders. These include the Swedish Food Agency, the Swedish Agricultural Board, local mussel farmers and the tourism sector. ODF Sweden is already in contact with these stakeholders and end users, and together we are forming a network to specify the exact needs for a HABs prognosis tool. The work in this project will cover the whole chain from research and development to product innovation and marine management through the use of ocean data in the digital blue economy.

The Killer Algae Challenge is currently ongoing. If you are interested in learning more or contributing to this challenge, please follow Algal Blooms Sweden on Facebook, and please contact us at email@oceandatafactory.se. Here is some background info.

Second cycle - The Koster challenge

Our second cycle (starting in January 2020) takes a dive into the Kosterhavet National Park, one of Sweden’s most important and unique marine environments. The Park is not only the first marine park to be established but also the most species-rich marine environment in all of Sweden. The area obtained official protected status in 2009 and monitoring changes in the marine environment has become a top national priority. Understanding this complex ecosystem and its development in light of a warming planet and increased human activity is crucial to ensuring its survival for generations to come.

Researchers studying the changes in the marine ecosystem over the last 30 years at Kosterhavet National Park face many challenges, including:

  • Storing and accessing observation data in a centralised and standardised way
  • Identifying species when most of the captured footage contains little to no fauna/flora
  • Analysing past footage with poor visibility conditions and/or low camera resolution

In the past, this footage has been examined and annotated manually by experts in marine biology. Given the advances in data science, this is simply not an efficient use of their time, and we believe that machine learning techniques could help to automate the parts of the process that significantly slow things down.

Thus, for its second innovation challenge, ODF Sweden decided to investigate the following question: Could we set up the data infrastructure for a highly-performant object detection model to help us detect an important habitat-building marine species (Lophelia Pertusa) in real-time footage?

This second innovation challenge has now been concluded, and our process and findings are documented in a blogpost written by Combine, one of ODF Sweden’s partners. In addition, there are links to our open source code repositories.

Central to the successful completion of this cycle was the open challenge/call for contributions in the Koster Seafloor Observatory on Zooniverse. 2493 volunteers!

First cycle - The Killer Shrimp Challange

Our first cycle (starting in July 2019) deals with an invasive species that has already caused considerable damage in Europe is the Killer Shrimp or Dikerogammarus Villosus. The presence of the killer shrimp has been recorded in rivers in Western Europe, presumably by travelling through inland waterways from the Black Sea and assumed to be carried by cargo ships where ocean expanses are too vast to traverse. However, its presence is yet to be detected in the three archipelagos of Stockholm, Åland and Turko, all in the Baltic Sea. As this shrimp is devastating for the local ecosystems that it invades, the questions to be investigated by ODF Sweden in collaboration with the Swedish Agency for Marine and Water Management (SwAM) included the following:

1) What are the factors that could lead to the spread of the shrimp into the Baltic? 2) How might these factors be effected by various scenarios such as changes in climate or shipping routes? 3) How might the spread of this species into the Baltic effect local ecosystems and even potentially local industry?

ODF Sweden has worked on a principle of openness in this first innovation challenge, using open datasets including the following:

1. Port locations in Europe (EMODNET Human Activities)
2. Ocean surface temperatures and salinity for Baltic Sea (SMHI) and North Sea regions (SeaDataNet)
3. Presence data of D. Villosus from observations ranging from 1928-2019 (GBIF)
4. Marine data layers (Bio-Oracle)
5. Ocean temperature and salinity (Marine Copernicus)

The first innovation challenge was executed as a deliverable within the project Three Archipelagos, led by SwAM. It has now been concluded and our process and findings are documented in a blogpost written by Combine, one of ODF Sweden’s partners.

ODF Sweden has made the methodology developed in this first use case available for external actors with the hope that it can be further developed and applied in other regions and for other species.

In addition, we have created an open challenge on Kaggle based on this first use case: The Killer Shrimp Challenge