Co-written by Nepal Red Cross, Cameroon Red Cross, and IFRC Solferino Academy

The Red Cross Red Crescent works in communities around the world. We asked – How can we better co-design and prototype emerging technology with communities at the center? Over the past year, Nepal Red Cross, Cameroon Red Cross, Nesta UK, and IFRC Solferino Academy embarked on this very journey. The teams explored how emerging technology, specifically Artificial Intelligence (AI), could support the effective delivery of humanitarian work by involving communities in the design in two very different contexts. This project was done during a global pandemic and ongoing emergencies with the everyday complexities of humanitarian response. 

Collective Crisis Intelligence methods combine the collective intelligence of local community actors (e.g. volunteers in local branches) with a wide range of additional data sources, artificial intelligence and predictive analytics to support crisis management and reduce the devastating impacts of humanitarian crises. (See our article: Introducing Collective Intelligence and the Nesta research.)

The Collective Crisis Intelligence project is one of the first attempts to develop and test new methods for involving crisis-affected populations and frontline responders in the development, evaluation and utilisation of new predictive AI models.

Register and join the upcoming event

Join us online on Tuesday, July 19th to hear more from the Cameroon and Nepal Red Cross Societies!

> Click here for Registration

Nepal Red Cross – “Non-Food Relief Items Predict” prototype

Nepal Red Cross conducted a CCI project in the area of Non-Food Relief Items (NFRI). They hosted a number of discussions and workshops – virtual and in-person to learn and co-design with local communities. Over 1000 people participated in workshops in both headquarters and in local districts. The team hosted four workshops (2 virtual and 2 in-person) with the Nepal Red Cross Society (NRSSC) Headquarters, NFRI team, and three workshops in 3 different districts (at the mountain, the hill, and the plain).

Collective intelligence is the sharing of knowledge and ideas with the help of people, data and scientific tools. The CCI model can be applied in other projects to grasp the needs of the community. The project teams within NRCS can be made aware to design the project from the researching point of view that best addresses the community’s needs. By and large, the recommendations of the project shall be adopted by the NRCS and be incorporated into the forthcoming ‘NFRI Manual’ to strengthen and standardise the NFRI system of NRCS.

This innovation project convened teams to learn and think more about artificial intelligence (AI). The project methodology was new to teams and helped them to talk more about the support to community people during emergencies. In a nutshell, the product of this research will help NRCS to predict different groups of NFRI and it may contribute to cost-effectiveness and service efficiency. NRCS is interested in developing a technology-based tool for blood donors and requirement prediction through a research process of all blood centres by analysing the demand and supply. 

Cameroon Red Cross Society – “Report and Respond” Prototype

The Cameroon Red Cross CCI project tackled rumors and misinformation by detailing how to sensitise and collect community feedback. The National Society, concerned with the well-being of the communities, is reflecting on how to capture and find answers to rumours in an effective and timely manner.

Going forward, Cameroon Red Cross is working to integrate the CCI methodology into the various projects as they see it as an efficient and effective tool for raising community awareness. Future decisions could make crisis intelligence an indispensable tool in the implementation of projects. In view of the expectations founded by the above-mentioned reflection, developing such a project would enable the National Society to offer a competitive tool for capturing and responding to rumours through digital solutions.

The team also took a network-based approach to involve people beyond the project to foster shared ownership. During the workshops in Yaoundé, staff from departments outside of the Communications department were involved and trained. The Cameroon Red Cross opted for a train-the-trainer model, in which they trained teams to train volunteers. Initially, the capacity-building strategy of training volunteers directly was successful in getting CRC staff to receive training from the trainers and to contribute to the training of the volunteers, which allowed the volunteers to take greater ownership of the training. For the CRC staff, this not only enabled them to supervise the volunteers with ease, but also to understand the project’s challenges during the project and envisage its sustainability.   

Cameroon Red Cross is currently exploring how to share specific information about this new technology with our communities and local stakeholders. When using emerging technology in our humanitarian work, there needs to be a common language and understanding to bridge the digital divide.  

Some key lessons from our teams

• The degree of volunteer engagement is important to success. This shapes the development – combining feedback to improve the functioning of the tool. 
• Community values and contexts play a large role in shaping any technology. This comes with ethical challenges and dilemmas to protect vulnerable communities and adhere to humanitarian values. 
• The innovation methodology gave National Societies the opportunity to learn new methods for working from project methodology/modalities, explore computer models of Artificial Intelligence (AI)  with community engagement, and be creative managing constraints and challenges. 
• While the AI models can not be reused across all types of engagement, the process/learning can be reused. 
• The innovative process needs local community engagement, however, feedback loops are not often possible in normal conversations. Teams need to plan for safety and alternative ways to get input. 
• There needs to be the translation of tools and processes /methodology to bridge the divide and foster organizational learning. 
• The digital divide and data infrastructure are barriers to realizing the full potential of emerging technology embedded in our work. 

Next steps

We can listen to the community better with modern and digital tools.

Hyacinthe Olinga Eloundou, Director of Communications and of the Promotion of humanitarian values, Cameroon Red Cross. 

While it’s clear that there are many interesting technical applications of AI that can improve efficiency in our humanitarian work, the purpose of AI for the humanitarian sector has to be about crisis-affected people being protected and honoured in their right to a high-quality response (vs improvement of international actors/funding flows/objectives).  We think that responsible and participatory AI in our sector will come down to adapting, creating and supporting good local governance systems. For these systems to be integral to the places where the data will be collected from, and hence data and identity vulnerability exists, the focus of the RCRC could be around supporting demonstrations where collaborations of local institutions and local (civic) organisations make sense practically of the new set of realities and responsibilities that AI brings. The option to create highly participatory AI systems gives room to these demonstrations being able to reshape dynamics between (crisis-affected) people, (civic) organisations and institutions. The Red Cross Red Crescent will be well-positioned to facilitate innovation in this interface thanks to the unique identity of being community-led and auxiliary to government.

Through this project, colleagues in Nepal and Cameroon identified and made sense of the implications of data technology on their operations and organisation, and this has a bigger meaning for the organisation at large. It’s therefore crucial that some actors in our organisation pioneer and test the boundaries of these new developments, as we facilitate collective learning to make sure we benefit from our networked model. 

Some questions to explore:

• How do we position ourselves so we are at the forefront of the development of deeply ethical AI? 
• How do we manage risks around data storage and humanitarian identity? 
• Who is the ‘owner’ of the data and the data processes? (e.g. data controller)? 
• How do we explore emerging technologies and simultaneously address the digital divide in our work?  

Even more than the learning about tool development, and its impact on the operations of National Societies, this project has given us the chance to start having these critical discussions as a global organisation. If you are interested in exploring more about these projects, please do get in touch!  [email protected] 

The Solferino Academy would like to thank the partners in this project for collaborating on such an ambitious, and often challenging, initiative. Our thanks goes out to the Cameroon Red Cross Team (including the amazing CEA fellow), Nepal Red Cross Society Team, Nesta UK (including the brilliant data science fellows), UK Humanitarian Innovation Hub , OpenLab (NewCastle University) and the many IFRC (regional) teams and colleagues that helped out and stepped up in various ways along the way.

Illustration: Technograms by ©Cartoonbase