CARE-KNOW-DO AI forest fire SDG15 Life on Land

AI for Protecting Life on Land (SDG 15) with the CARE-KNOW-DO Framework

1. AI for Protecting Life on Land (SDG 15) with the CARE-KNOW-DO Framework Nektarios Kokolakis Med Computer Science and Economics This CARE-KNOW-DO science-action was supported by G. Panselinas and A. Okada

2. In compliance with

3. What is open schooling supported by CARE-KNOW-DO? Real-life issue that students care about for sustainable development goals Curriculum knowledge in context supporting students’ citizenship and future career Fun participatory approaches for learning in action Students interactions with families, teachers , and professionals (Okada, 2023)

4. Students interactions with families, teachers, and AI professionals Community engagement (Okada, 2023)

5. CARE KNOW DO Agency Accountability Responsability Ethics Safety Fairness Value Human-centre AI for Real-world Challenge Understad AI for Digital Lifelong Learning Apply AI for Inclusive Sustainable Future CARE-KNOW-DO principles to support AI in Education (Okada, 2024) AI in Education for SDG

6. AI Competency framework for Teachers for Students (Okada, 2024) AI in Education for SDG

7. Machine Learning and Image Recognition in the service of the environment Nektarios Kokolakis Med Computer Science and Economics This CARE-KNOW-DO science-action was supported by G. Panselinas and A. Okada

8. Real Issue in Greece: Forest Fire Climate Challenges in Greece Greece is increasingly facing severe climate challenges. These include more frequent and intense heatwaves, droughts, and wildfires, which are exacerbated by climate change. The Mediterranean region, where Greece is located, is particularly vulnerable to these effects. The summer of 2021 saw catastrophic wildfires that ravaged large parts of the country, destroying forests, homes, and livelihoods. Importance of Weather Forecasting Accurate weather forecasting is crucial for several reasons: Disaster Preparedness : Timely and accurate weather forecasts can help prepare for and mitigate the impact of extreme weather events like heatwaves, storms, and wildfires. Agriculture : Farmers rely on weather forecasts to make informed decisions about planting, irrigation, and harvesting. Public Safety : Forecasts help protect public health by warning of extreme temperatures and air quality issues. Tourism : Greece's economy heavily depends on tourism, which can be significantly impacted by weather conditions. Reliable forecasts can enhance tourist safety and experience.

9. The integration of AI in weather forecasting The use of machine learning for recognising cloud formations, offers significant potential for improving the accuracy and reliability of weather predictions in Greece. This is vital for mitigating the impacts of forest fires, protecting natural resources, and ensuring public safety. With climate change expected to exacerbate the frequency and severity of wildfires, advanced forecasting tools will be essential in the country's efforts to manage and respond to these threats effectively. Recent Incidents 2018 Mati Fire : Killed over 100 people, caused extensive property damage, and spread rapidly due to strong winds and high temperatures. 2021 Wildfires : Burned thousands of hectares, destroyed homes, and led to mass evacuations, fueled by a prolonged heatwave and dry conditions. https:// en.wikipedia.org /wiki/2018_Attica_wildfires https:// www.theguardian.com /world/2021/ aug /08/ thousands-flee-greek-island-evia-wildfires-raze-forest-and-homes 2021 Greece Wildfires

10. Importance of forest protection

11. Classification System

12. Classification System

13. Deforestation risk practical examples

14. Science-action in schools

15. Supporting Non-formal learning (Students with AI Professionals )

16. Discuss the following topics. We emphasize both the social and ethical and environmental dimensions. The discussion will take place at school and continue at home with their parents . You choose as many as you want or as many as you can manage. You can also use your own themes. Duration : 2 teaching hours at school. The first will include discussion and video viewing and the second will include the following deliverable and the first preparation of questions for the scientist. Deliverables : The children in the second teaching hour will be divided into two groups and will write in the word processor their own conclusions about the usefulness of image recognition, its positives and negatives. They will also write the first questions for the scientist. Note Take detailed notes of all that you discuss with your parents, so that we can use them in the Classroom. Activities

17. Topics - Face and Image Recognition (many smart phones now unlock by recognising the face of their owner) - Google search by face - Autonomous car driving - Using image recognition to make our cities safer. - Using image recognition to protect the environment Concerns - Where the data generated by facial recognition will be stored 2 - Who will manage it . For what purposes? Supporting texts - The problem of bias in facial recognition - Advantages and disadvantages of facial recognition - Usefulness of image recognition - Harnessing Artificial Intelligence for the Earth Note Take detailed notes of all that you discuss with your parents, so that we can use them in the Classroom. Activities

18. Activities - Other countries using the smart city model and, in particular, facial recognition, are The UK, the US, China, India, South Korea and others. The smart city model certainly has its place in a technologically advanced society. The essential question, however, is whether it is compatible with a democratic society. The security it promises is important. But is it really necessary and more effective than human actions? And above all, is it ethical towards the human being and his individual freedoms? - In September 2019, four researchers wrote to the publisher Wiley"respectfully " to immediately retract a scientific paper. The study, published in 2018 4 , had trained algorithms to distinguish the faces of Uyghurs (a predominantly Muslim minority ethnic group in China) from those of Korean and Tibetan ethnicity. China had already been condemned internationally for its heavy surveillance and mass detention of Uighurs in camps in the northwestern province of Xinjiang. According to media reports, the authorities in Xinjiang used surveillance cameras equipped with software tailored to the faces of the Uighurs. As a result, many researchers found it disturbing that academics had tried to create such algorithms - and that an American journal had published a research paper on the subject. - Machine learning, through image recognition, can categorize animals based on images alone. When environmentalists or wildlife experts capture images or photographs, they can ask machine learning systems to process the data and accurately classify animals. This step helps monitor populations and behaviours for appropriate conservation. Placing cameras with facial recognition capability in public places, the city is able to "track" its residents. The aim of the project is to improve citizen safety and law enforcement. However, beyond the security that such a system is supposed to provide to citizens, it also threatens their privacy. Citizens cannot move freely while their personal data is collected with the possibility of being used for unknown purposes, legitimate or illegitimate. For example, in 2015, scientists at Stanford University in California published a set of 12,000 images from a webcam in a San Francisco café that had been broadcast live on the internet2. The following year, researchers at Duke University in Durham, North Carolina, released more than 2 million video frames (85 minutes) of footage of students walking around campus3 AI F acial recognition capability in public places

19. AI algorithm for rainforest deforestation risk model PrevisIA can identify informal roads, simulating future scenarios to stop rainforest loss events, such as forest fires, before they happen. R isk maps, probability maps and control matrices will not only allow Imazon to predict future deforestation, but will also create hazard alerts and use artificial intelligence to cover more areas at scale. It has been designed with policy makers in mind so they can make better decisions to protect our biodiversity, and ecologists can use the same toolkits to support their mission. The model will be more accurate and allow for expanded data inputs.

20. 1 https://www.nature.com/articles/d41586-020-03187-3#ref-CR1 2 https://exposing.ai/datasets/ 3 https://arxiv.org/abs/1609.01775 4 https://wires.onlinelibrary.wiley.com/doi/epdf/10.1002/widm.1278 5 https://waymo.com/waymo-driver/ 6 https://jivp-eurasipjournals.springeropen.com/articles/10.1186/s13640-018-0391-6 7 https://www.microsoft.com/en-us/ai/ai-for-earth-imazon References KNOW  Videos in Greek selected by teachers and recommended by researchers https:// www.youtube.com / watch?v =Uk6p2eGfLtg  https:// www.youtube.com / watch?v =0dKkMS5-0jA  https:// www.youtube.com / watch?v =67ik8tnxqlE  https:// www.youtube.com / watch?v =eo5ugIihZio  https:// www.youtube.com / watch?v =3ImwGReO3dg 

21. Supporting informal learning (students with their families)

22. KNOW Activity sheet for discussion at home with parents After informing your parents about what we discussed at school, discuss the following topics with them: 1. Ask them if they know of any examples of image recognition technology from their daily lives. 2. Tell you if, through these examples, they find it useful and appropriate to use this technology. 3. Help them to think of examples where this technology could be used to help solve serious problems, for example, environmental pollution, poverty, saving forests, reducing inequalities, sustainable development, etc. 4. Do the following activity with them.

23. KNOW Activity sheet for discussion at home with parents - Read together with your parents the following text: Imazon , along with Fundo Vale and Microsoft are integrating an existing rainforest deforestation risk model algorithm into Azure. With a more robust forest image processing model, PrevisIA can identify informal roads, one of the leading indicators of future deforestation, simulating future scenarios to stop rainforest loss events, such as forest fires, before they happen. The resulting risk maps, probability maps and control matrices will not only allow Imazon to predict future deforestation, but will also create hazard alerts and use artificial intelligence to cover more areas at scale. The model will be more accurate and allow for expanded data inputs. It has been designed with policy makers in mind, so they can make better decisions to protect our biodiversity, and ecologists can use the same toolkits to support their mission. - Could something similar be used in Greece, or even better in your region? If so how and for what reasons? Note Take detailed notes of all that you discuss with your parents, so that we can use them in the Classroom.

24. Supporting formal learning ( students with school teachers)

25. Sustainable Development Goals 13. Climate Action 15. Life on Land 16. Promotion of Peaceful and Inclusive Societies Curriculum Knowledge: Programming, Internet, Multimedia, Internet Security Skills: Programming, Image Processing, AI awareness Attitude & Values: Environmental Protection, Personal Data, Human Rights, Teamwork Connection of Curriculum – Scenario The scenario can be implemented in all grades of middle school within the context of the Informatics course. If taught in the first grade, it should follow a few introductory lessons in Scratch. CARE-KNOW-DO Unit Care: Internet, Search Engines, Internet Security Unit Know: Image Processing, Programming Unit Do: Programming This Photo by Unknown Author is licensed under CC BY

26. Instructions about Machine Learning for teachers to support students Creating a Managed Classroom Account on Machine Learning for Kids Attention: This scenario only supports managed accounts. If you want to use unmanaged accounts, you do so at your own risk. To create a managed classroom account, go to the link https://machinelearningforkids.co.uk/#!/login , click Sign Up, select Teacher, and read the option for creating a managed classroom account to see what you need to do. Creating Students From the Teacher menu, select Manage Students. Then create a group of students. After that, add students to the group. It is better to add groups of students rather than individual students. Declare the username and the password is created automatically. Loading Scratch Applications The Scratch applications for the Know and Do activities can only be opened through Machine Learning for Kids by selecting implement, opening Scratch from there, and then choosing load from Scratch. They do not open from the regular Scratch application. This Photo by Unknown Author is licensed under CC BY

27. Scratch is a visual programming language designed for children and beginners, developed by the MIT Media Lab. It allows users to create interactive stories, games, and animations by snapping together blocks of code, making programming accessible and easy to understand. With its highly visual interface and block-based coding system, Scratch eliminates the need for typing complex syntax, enabling users to focus on learning core programming concepts. Additionally, Scratch has a vibrant online community where users can share their projects and collaborate, fostering creativity and problem-solving skills. Scratch is widely used in educational settings, providing a fun and engaging way to introduce students to the world of coding. This Photo by Unknown Author is licensed under CC BY-SA

28. Research Analysis about students ‘ learning with AI with CONNECT-science self-reported instrument Okada and Panselinas (2023) CONNECT-science instrument

29. STUDENTS PARTICIPANTS Pilot study: 159 students

30. Students’ views about their learning with AI Μάθ α με γι α το π ερι β άλλον , Έμ α θ α νέ α π ράγμ α τ α γι α την ε π ιστήμη . Μάθ α με π ολλά κ α ινούργι α π ράγμ α τ α π άνω στο θέμ α της Τεχνητής νοημοσύνη . Μάθ α με γι α την α ν α γνώριση εικον α ς . ενημερώθηκε η οικογένει α γι α την π ως λειτουργάει η Τεχνητής νοημοσύνη Μάθ α με γι α Πώς οι γονείς μου β λέ π ουν την ε π ιστήμη We learned about the environment, I learned new things about science. We learned a lot of new things on the subject of Artificial Intelligence. We learned about image recognition. Our family was informed about how Artificial Intelligence works. We learned about how our parents think about science. Greek Students

31. Students’ views about their learning with AI

32. Students’ views about their learning with AI

33. Enjoyment and Engagement: 84% who want to participate in new science activities and 54% of students find learning science fun. This suggests that even some students who don't necessarily find science fun are still interested in hands-on experiences. Enjoyment to Career Interest: students who would like a job that uses science (48%) is close to those who find learning science fun (54%). This suggests a strong correlation between enjoying science and considering it for a career. It indicates that enjoyment of science is indeed important for career aspirations in the field. Fun vs. Expertise : While 54% find learning science fun, only 31% want to be seen as experts. This might indicate that while students enjoy science they are not sure about science as part of their identity. Activity Participation vs. Career Interest: The very high interest in participating in science activities (84%) compared to wanting a science-related job (48%) suggests that practical, engaging science experiences are much more appealing than the prospect of a science career. Students’ views about their learning with AI Full Report [ URL ]

34. REFERENCES Okada, A. (Ed.). (2023). Inclusive open schooling with engaging and future-oriented science: Evidence-based practices, principles & tools. Milton Keynes, UK: The Open University. Okada, A.(2023). Knowledge Cartography for Young Thinkers: Sustainability Issues, Mapping Techniques and AI Tools. Advanced Information and Knowledge Processing. Switzerland: Springer (In Press). Okada, A.; Sherborne T; Pandelinas G. Kolionis G. (2024, under review). AI in Education: A Cross-National Study of Open Schooling Using the CARE-KNOW-DO Framework for Sustainable Development Goals. Panselinas , G. (2023). Operationalising Open schooling on Scale for Science and Sustainability Curricula: The case of Greece. Proceedings of CICOS Conference, Barcelona, 4-5 July 2024 DOI:10.5281/zenodo.1014901