Periodic Reporting for period 2 - CROWD4SDG (Citizen Science for Monitoring Climate Impacts and Achieving Climate Resilience)
Reporting period: 2021-08-01 to 2023-04-30
Problem context: The 17 Sustainable Development Goals (SDGs), launched by the UN in 2015, are underpinned by hundreds of concrete targets and indicators. However, for many of these targets and indicators, National Statistical Offices (NSOs) do not have the data collection capacity. Measuring progress towards the SDGs is thus a challenge that can benefit from non-standard data sources, such as Citizen Science (CS).
Importance for Society: The goal of the Crowd4SDG project was to research the extent to which CS can provide an essential source of non-traditional data for tracking progress towards the SDGs, as well as the ability of CS to generate social innovations that enable such progress. Based on shared expertise in crowdsourcing for disaster response, the transdisciplinary Crowd4SDG consortium of six partners focused on SDG 13, Climate Action, to explore new ways of applying CS for monitoring the impacts of extreme climate events and strengthening the resilience of communities to climate-related disasters.
Project Objectives: To achieve this goal, Crowd4SDG initiated research on the applications of artificial intelligence and machine learning to enhance CS and explored the use of social media and other non-traditional data sources for more effective monitoring of SDGs by citizens. Crowd4SDG used direct channels through consortium partner UNITAR to provide National Statistical Offices (NSOs) with recommendations on best practices for generating CS data for tracking the SDGs.
To this end, Crowd4SDG rigorously assessed the quality of the scientific knowledge and usefulness of practical innovations occurring when teams develop new CS projects focusing on climate action. This occured through three annual challenge-based innovation events, involving online coaching. A wide range of stakeholders, from the UN, governments, the private sector, NGOs, academia, innovation incubators and maker spaces were involved in advising the project and exploiting the scientific knowledge and technical innovations that it generated.
At the core of the project was a novel innovation cycle called GEAR (Gather, Evaluate, Accelerate, Refine), which ran once a year. The GEAR cycles involved online selection and coaching of citizen-generated ideas for climate action, using the UNIGE Open Seventeen Challenge (O17). The most promising projects were accelerated during an online two-month Challenge-Based Innovation (CBI) workshop. Top projects received further support at an annual SDG conference, the Geneva Trialogue.
Each Crowd4SDG GEAR cycle focused on a specific aspect of Climate Action and its connection with other SDGs. The first GEAR cycle was on Climate Change and its impact on Urban Water Resilience (SDG 6 and 11). The second GEAR cycle was on Climate Change and Gender Equality (SDG 5). The third GEAR cycle was on Climate Adaptation and Social Justice (SDG 16).
Importance for Society: The goal of the Crowd4SDG project was to research the extent to which CS can provide an essential source of non-traditional data for tracking progress towards the SDGs, as well as the ability of CS to generate social innovations that enable such progress. Based on shared expertise in crowdsourcing for disaster response, the transdisciplinary Crowd4SDG consortium of six partners focused on SDG 13, Climate Action, to explore new ways of applying CS for monitoring the impacts of extreme climate events and strengthening the resilience of communities to climate-related disasters.
Project Objectives: To achieve this goal, Crowd4SDG initiated research on the applications of artificial intelligence and machine learning to enhance CS and explored the use of social media and other non-traditional data sources for more effective monitoring of SDGs by citizens. Crowd4SDG used direct channels through consortium partner UNITAR to provide National Statistical Offices (NSOs) with recommendations on best practices for generating CS data for tracking the SDGs.
To this end, Crowd4SDG rigorously assessed the quality of the scientific knowledge and usefulness of practical innovations occurring when teams develop new CS projects focusing on climate action. This occured through three annual challenge-based innovation events, involving online coaching. A wide range of stakeholders, from the UN, governments, the private sector, NGOs, academia, innovation incubators and maker spaces were involved in advising the project and exploiting the scientific knowledge and technical innovations that it generated.
At the core of the project was a novel innovation cycle called GEAR (Gather, Evaluate, Accelerate, Refine), which ran once a year. The GEAR cycles involved online selection and coaching of citizen-generated ideas for climate action, using the UNIGE Open Seventeen Challenge (O17). The most promising projects were accelerated during an online two-month Challenge-Based Innovation (CBI) workshop. Top projects received further support at an annual SDG conference, the Geneva Trialogue.
Each Crowd4SDG GEAR cycle focused on a specific aspect of Climate Action and its connection with other SDGs. The first GEAR cycle was on Climate Change and its impact on Urban Water Resilience (SDG 6 and 11). The second GEAR cycle was on Climate Change and Gender Equality (SDG 5). The third GEAR cycle was on Climate Adaptation and Social Justice (SDG 16).
As a research highlight, Crowd4SDG designed and implemented a new tool called VisualCit, that used AI to rapidly filter large social media data sets to find the most relevant images for further processing by crowdsourcing, using a platform called Citizen Science Project Builder (CSPB), also developed by the Crowd4SDG consortium.
As part of Crowd4SDG’s response to a request from the Commission to contribute to pandemic-related research, VisualCit was used for automatically analysing over 3.5M images from social media about wearing masks during the pandemic, and filtering out 5k images that required further detailed classification using CSPB. The result provided data about the adoption of policies for wearing masks in public places during the Covid-19 pandemic. Data generated by this combination of AI and crowdsourcing showed correlation with information gathered from governmental institutions using traditional surveys, and the results have been published in a major international peer-reviewed conference.
As an innovation highlight, Crowd4SDG coached some 50 youth-led SDG innovation projects, representing 150 young innovators. Nine of these projects were selected for insertion in an SDG Accelerator programme at the University of Geneva, two of them were subsequently selected for support by local social incubators, and two of them were supported by a follow-on project called Yoma led by UNICEF, supported by the Fondation Botnar.
As part of Crowd4SDG’s response to a request from the Commission to contribute to pandemic-related research, VisualCit was used for automatically analysing over 3.5M images from social media about wearing masks during the pandemic, and filtering out 5k images that required further detailed classification using CSPB. The result provided data about the adoption of policies for wearing masks in public places during the Covid-19 pandemic. Data generated by this combination of AI and crowdsourcing showed correlation with information gathered from governmental institutions using traditional surveys, and the results have been published in a major international peer-reviewed conference.
As an innovation highlight, Crowd4SDG coached some 50 youth-led SDG innovation projects, representing 150 young innovators. Nine of these projects were selected for insertion in an SDG Accelerator programme at the University of Geneva, two of them were subsequently selected for support by local social incubators, and two of them were supported by a follow-on project called Yoma led by UNICEF, supported by the Fondation Botnar.
By organizing events with a range of stakeholders in UN agencies, International Organizations and National Statistical Offices, the Crowd4SDG partners were able to curate a number of specific challenges for future GEAR cycles. This process enabled the project to generate data for SDG indicators that was useful for the SDGs, as confirmed by NSOs.
Based on the results from the first GEAR Cycle, the partners put a greater focus on several key issues, to increase the socio-economic impact of the project : ensuring working prototypes were generated during the second and third GEAR cycles, assisting teams to reach out to suitable communities to support the CS projects generated during each GEAR cycle, and ensuring that the teams identify suitable local incubators at an early stage, to nurture their projects after each GEAR cycle.
Consortium partners will maintain some of the Crowd4SDG AI-enhanced CS tools so that these can be used in future SDG projects. Consortium partners have also developed tools for collecting digital traces generated during the innovation process, which are being used to measure - and ultimately to increase - diversity, activity, and interactions of participants in the follow-on Yoma project led by UNICEF.
Based on the results from the first GEAR Cycle, the partners put a greater focus on several key issues, to increase the socio-economic impact of the project : ensuring working prototypes were generated during the second and third GEAR cycles, assisting teams to reach out to suitable communities to support the CS projects generated during each GEAR cycle, and ensuring that the teams identify suitable local incubators at an early stage, to nurture their projects after each GEAR cycle.
Consortium partners will maintain some of the Crowd4SDG AI-enhanced CS tools so that these can be used in future SDG projects. Consortium partners have also developed tools for collecting digital traces generated during the innovation process, which are being used to measure - and ultimately to increase - diversity, activity, and interactions of participants in the follow-on Yoma project led by UNICEF.