- data quality, by simplifying expert verification, or using AI to improve identification;
- data interoperability, by facilitating integration between data sets;
- data accessibility, by processing data so they can be more easily uploaded to the European Open Science Cloud  ensuring open access;
- the sustainability and maintenance of citizen observatories, for example, through low-cost tools that can be used, adapted and replicated.
- by capitalising on the interoperability of data, which will lead to larger, integrated and more useful data sets;
- by building on co-design approaches, involving and combining the citizens’ and researchers’ experience and expertise throughout the project’s life cycle;
- by feeding and supporting the education and learning experience of building on, participating in and contributing to sustainable development and global citizenship (which is the focus of SDG 4.7, as well as a cross-cutting theme across all SDGs).
2. Materials and Methods
2.1. Citizen Observatories: Opportunities to Support the SDGs
2.2. Limitations of the Citizen Observatory Method
2.3. The Cos4Cloud Methodology for Improving Citizen Observatory Networks
3.1. How Cos4Cloud Improves Interoperability: AI-Based Plant Identification through [email protected]
3.1.1. [email protected] Identification as a Service
3.1.2. [email protected] Research-Grade Data
3.2. How Cos4Cloud Enables Citizen Observatories to Support Education and Learning
4.1. Citizen Observatory Networks for Society
4.2. Citizen Observatory Networks for Policy
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
- United Nations. Resolution A/RES/70/1. Transforming Our World: The 2030 Agenda for Sustainable Development. In Seventieth United Nations General Assembly; United Nations: New York, NY, USA, 2015; Available online: www.un.org/ga/search/view_doc.asp?symbol=A/RES/70/1&Lang=E (accessed on 27 February 2022).
- Fraisl, D.; Campbell, J.; See, L.; Wehn, U.; Wardlaw, J.; Gold, M.; Moorthy, I.; Arias, R.; Piera, J.; Oliver, J.L.; et al. Mapping citizen science contributions to the UN sustainable development goals. Sustain. Sci. 2020, 15, 1735–1751. [Google Scholar] [CrossRef]
- Fritz, S.; See, L.; Carlson, T.; Haklay, M.M.; Oliver, J.L.; Fraisl, D.; Mondardini, R.; Brocklehurst, M.; Shanley, L.A.; Schade, S.; et al. Citizen science and the United Nations Sustainable Development Goals. Nat. Sustain. 2019, 2, 922–930. [Google Scholar] [CrossRef]
- Ceccaroni, L.; Bowser, A.; Brenton, P. Civic education and citizen science: Definitions, categories, knowledge representation. In Analyzing the Role of Citizen Science in Modern Research; Ceccaroni, L., Piera, J., Eds.; IGI Global: Hershey, PA, USA, 2017; pp. 1–23. ISBN 9781522509622. [Google Scholar]
- Shulla, K.; Filho, W.L.; Sommer, J.H.; Salvia, A.L.; Borgemeister, C. Channels of collaboration for citizen science and the sustainable development goals. J. Clean. Prod. 2020, 264, 121735. [Google Scholar] [CrossRef]
- Moczek, N.; Voigt-Heucke, S.; Mortega, K.; Cartas, C.F.; Knobloch, J. A Self-Assessment of European Citizen Science Projects on Their Contribution to the UN Sustainable Development Goals (SDGs). Sustainability 2021, 13, 1774. [Google Scholar] [CrossRef]
- Schleicher, K.; Schmidt, C. Citizen Science in Germany as Research and Sustainability Education: Analysis of the Main Forms and Foci and Its Relation to the Sustainable Development Goals. Sustainability 2020, 12, 6044. [Google Scholar] [CrossRef]
- Sprinks, J.; Woods, S.; Parkinson, S.; Wehn, U.; Joyce, H.; Ceccaroni, L.; Gharesifard, M. Coordinator Perceptions When Assessing the Impact of Citizen Science towards Sustainable Development Goals. Sustainability 2021, 13, 2377. [Google Scholar] [CrossRef]
- FreshWater Watch. Available online: https://freshwaterwatch.thewaterhub.org/ (accessed on 27 February 2022).
- Hegarty, S.; Hayes, A.; Regan, F.; Bishop, I.; Clinton, R. Using citizen science to understand river water quality while filling data gaps to meet United Nations Sustainable Development Goal 6 objectives. Sci. Total Environ. 2021, 783, 146953. [Google Scholar] [CrossRef]
- UNEP. Measuring Progress: Environment and the SDGs; United Nations Environment Programme: Nairobi, Kenya, 2021. [Google Scholar]
- UNEP. Global Environment Outlook—GEO-6: Healthy Planet, Healthy People; United Nations Environment Programme: Nairobi, Kenya, 2019; ISBN 978-1-108-70766-4. [Google Scholar]
- MICS. Available online: https://mics.tools/ (accessed on 27 February 2022).
- Cos4Cloud. Available online: https://cos4cloud-eosc.eu/ (accessed on 27 February 2022).
- WeObserve. Citizen Observatories; International Institute for Applied Systems Analysis (IIASA): Laxembourg, 2018; Available online: www.weobserve.eu/about/citizen-observatories/ (accessed on 27 February 2022).
- Ceccaroni, L.; Piera, J.; Wernand, M.R.; Zielinski, O.; Busch, J.A.; Van der Woerd, H.J.; Bardaji, R.; Friedrichs, A.; Novoa, S.; Thijsse, P.; et al. Citclops: A next-generation sensor system for the monitoring of natural waters and a citizens’ observatory for the assessment of ecosystems’ status. PLoS ONE 2020, 15, e0230084. [Google Scholar] [CrossRef]
- ESOC Portal. Available online: https://eosc-portal.eu/ (accessed on 27 February 2022).
- Liu, H.-Y.; Kobernus, M.; Broday, D.; Bartonova, A. A conceptual approach to a citizens’ observatory—Supporting community-based environmental governance. Environ. Health 2014, 13, 107. [Google Scholar] [CrossRef]
- Rathnayake, C.; Joshi, S.; Cerratto-Pargman, T. Mapping the current landscape of citizen-driven environmental monitoring: A systematic literature review. Sustain. Sci. Pr. Policy 2020, 16, 326–334. [Google Scholar] [CrossRef]
- Ajates, R. Reducing the Risk of Co-Optation in Alternative Food Networks: Multi-Stakeholder Cooperatives, Social Capital, and Third Spaces of Cooperation. Sustainability 2021, 13, 11219. [Google Scholar] [CrossRef]
- Joly, A.; Bonnet, P.; Goëau, H.; Barbe, J.; Selmi, S.; Champ, J.; Dufour-Kowalski, S.; Affouard, A.; Carré, J.; Molino, J.-F.; et al. A look inside the [email protected] experience. Multimed. Syst. 2015, 22, 751–766. [Google Scholar] [CrossRef]
- Lorieul, T. Uncertainty in Predictions of Deep Learning Models for Fine-Grained Classification. Ph.D. Thesis, Université de Montpellier, Montpellier, France, 2020. [Google Scholar]
- Affouard, A.; Lombardo, J.-C.; Goeau, H.; Bonnet, P.; Joly, A. [email protected] App in the Era of Deep Learning. In Proceedings of the 5th International Conference on Learning Representations, ICLR 2017, Workshop Track Proceedings, Toulon, France, 24–26 April 2017; Available online: OpenReview.net (accessed on 27 February 2022).
- [email protected] Available online: https://my.plantnet.org/ (accessed on 27 February 2022).
- Affouard, A.; Joly, A.; Lombardo, J.-C.; Champ, J.; Goeau, H.; Bonnet, P. [email protected] Observations; Version 1.2. [email protected] Occurrence Dataset. 2020. Available online: https://www.gbif.org/dataset/7a3679ef-5582-4aaa-81f0-8c2545cafc81 (accessed on 1 January 2021).
- Sanders, E.B.-N.; Stappers, P.J. Co-creation and the new landscapes of design. CoDesign 2008, 4, 5–18. [Google Scholar] [CrossRef]
- Cos4Bio. Available online: https://cos4bio.eu/ (accessed on 27 February 2022).
- Short, P.C. Responsible Environmental Action: Its Role and Status in Environmental Education and Environmental Quality. J. Environ. Educ. 2009, 41, 7–21. [Google Scholar] [CrossRef]
- Jensen, B.B.; Schnack, K. The action competence approach in environmental education. Environ. Educ. Res. 2006, 12, 471–486. [Google Scholar] [CrossRef]
- Breiting, S.; Sorensen, F. Action competence and environmental education. Camb. J. Educ. 1999, 29, 349–353. [Google Scholar] [CrossRef]
- Citizen Science: Innovation in Open Science, Society and Policy. Available online: https://discovery.ucl.ac.uk/id/eprint/10058422/1/Citizen-Science.pdf (accessed on 27 February 2022).
- Tauginienė, L.; Butkevičienė, E.; Vohland, K.; Heinisch, B.; Daskolia, M.; Suškevičs, M.; Portela, M.; Balázs, B.; Prūse, B. Citizen science in the social sciences and humanities: The power of interdisciplinarity. Palgrave Commun. 2020, 6, 89. [Google Scholar] [CrossRef]
- Mominó, J.M.; Piera, J.; Jurado, E. Citizen Observatories as Advanced Learning Environments. In Analyzing the Role of Citizen Science in Modern Research; IGI Global: Hershey, PA, USA, 2017; pp. 192–212. [Google Scholar]
- UNESCO. Education for Sustainable Development: A Roadmap. 2020. Available online: https://unesdoc.unesco.org/ark:/48223/pf0000374802 (accessed on 27 February 2022).
- UNESCO. Framework for the Implementation of Education for Sustainable Development for the Period of 2020–2030. 2009. Available online: https://unesdoc.unesco.org/ark:/48223/pf0000370215 (accessed on 27 February 2022).
- UNESCO. Education for Sustainable Development Goals: Learning Objectives. Paris: UNESCO. 2017. Available online: https://unesdoc.unesco.org/ark:/48223/pf0000247444?posInSet=2&q (accessed on 27 February 2022).
- Bonney, R.; Shirk, J.L.; Phillips, T.B.; Wiggins, A.; Ballard, H.L.; Miller-Rushing, A.J.; Parrish, J.K. Next Steps for Citizen Science. Science 2014, 343, 1436–1437. [Google Scholar] [CrossRef]
- Jordan, R.C.; Gray, S.T.; Howe, D.V.; Brooks, W.R.; Ehrenfeld, J.G. Knowledge gain and behavioral change in citizen-science programs. Conserv. Biol. 2011, 25, 1148–1154. [Google Scholar] [CrossRef]
- Jordan, R.C.; Ballard, H.L.; Phillips, T. Key issues and new approaches for evaluating citizen-science learning outcomes. Front. Ecol. Environ. 2012, 10, 307–309. [Google Scholar] [CrossRef]
- Bonnet, P.; Joly, A.; Faton, J.; Brown, S.; Kimiti, D.; Deneu, B.; Servajean, M.; Affouard, A.; Lombardo, J.; Mary, L.; et al. How citizen scientists contribute to monitor protected areas thanks to automatic plant identification tools. Ecol. Solut. Évid. 2020, 1, e12023. [Google Scholar] [CrossRef]
- Schade, S.; Pelacho, M.; van Noordwijk, T.G.E.; Vohland, K.; Hecker, S.; Manzoni, M. Citizen Science and Policy. In The Science of Citizen Science; Vohland, K., Ed.; Springer: Berlin/Heidelberg, Germany, 2020; pp. 351–371. [Google Scholar] [CrossRef]
- Burgess, H.K.; DeBey, L.B.; Froehlich, H.E.; Schmidt, N.; Theobald, E.J.; Ettinger, A.K.; HilleRisLambers, J.; Tewksbury, J.; Parrish, J.K. The science of citizen science: Exploring barriers to use as a primary research tool. Biol. Conserv. 2017, 208, 113–120. [Google Scholar] [CrossRef]
|Set-valued species classification||The service returns a list of species corresponding to the query, each associated with a probability output by the deep learning model.||The probability output allows for better quality control in the downstream applications provided to end users.||An application may suggest to its end users to re-submit a better-quality photo if the confidence score returned for identification is too low. If the confidence score is very high, it can automatically record the observation for the end user.|
|Flora selection(optional)||This option allows the returned species to be filtered according to a given flora (among 41 possible floras at the time of writing).||The identification quality is improved, and the service can be more easily customised for specific use cases.||An application will not display a visually similar species for a given identification request (of the same genus, for example) if the species is not within the selected flora.|
|Similarity search||For each recognised species, the service returns the closest pictures to the query in the [email protected] database.||This allows better quality control of the identification by the end user. It also provides access to [email protected] images to illustrate the species.||Application developers do not have to manage visual data to illustrate the identified species and can benefit from the data shared by [email protected]|
|Multi-images and multi-organ observations||Several pictures of the same individual plant can be provided as input of the service (particularly photos of the different organs, including the flower, the fruit or the leaf).||The identification quality is improved, and the service can be more easily customised for specific use cases.||A biodiversity monitoring app can be used whatever the season or the phenological stage of the plant.|
|Key Features||Description||Added Value||Example|
|Regularly updated and enriched identification model||[email protected]’s AI model used for the identification service is regularly updated based on the collaboratively revised [email protected] database and research progress.||The identification quality is regularly improved, and the number of recognised species increases.||Newly identified species are rapidly recognisable by other users|
|Automated rejection of inappropriate content||Any picture of non-plant entities is automatically tagged as “Reject”.||Inappropriate content can be filtered before it enters the data flow of the downstream applications (database, app screens, etc.).||This feature facilitates the integration of the service into educational applications, ensuring the identification of only visual plant data.|
|API dashboard and documentation||The website my.plantnet.org provides rich documentation of the service as well as a dashboard allowing the users to follow its usage of the service.||Developers can easily integrate the service into downstream applications and monitor their usage.||An application using this service can monitor its consumption of the service in terms of number of queries per day, week, month, etc.|
|Quality of Service||Description||Added Value||Example|
|Non-regression tests of the identification quality||The quality of the identification is measured on several benchmarks before each model’s update.||The user has the guarantee that the identification quality is not degrading.||An application using this service can enrich the amount of information on each new identifiable species, these species being always identifiable in future versions of the recognition model.|
|Quality of service monitoring||The quality of service is monitored continuously based on several criteria (status, response time and scalability) and an alert system allows the developers to react quickly.||Service downtime is limited.||An application using this service can use it at any time.|
|Secured access with an API key||The use of the service requires a token provided at account creation (with email validation).||This highly limits the risk of attacks.||This limits the risk of denial of service attacks.|
|Scalability||[email protected]’s infrastructure and data management software are highly scalable.||Developers of downstream applications do not have to manage the workload themselves.||A third-party application can have a peak in usage or a large increase in users without causing the service to slow down.|
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).