Kenya: Conduct Research on Use of Artificial Intelligence (AI) in Anticipatory Action in Improving Wellbeing Outcomes for Children Affected by Climate Risks

ChildFund is an international child-centred development organization. We are a member of the ChildFund Alliance; a global network of 11 organizations that assist more than 36 million children in 70 countries around the world. In Kenya, ChildFund works through 13 local partners (LP’s) comprised of 50 community organizations in 27 counties. Barnfonden is a child rights organization based in Malmö, Sweden, working with children´s rights and safety in vulnerable areas that are heavily affected, or at risk of being affected by climate change. SVEP Design Center is an engineering and IT design innovation centre in Sweden that designs and specializes in project-based services for companies looking to develop cutting-edge tech products. Barnfonden through its relationship with SVEP has commissioned ChildFund to engage an external expert to undertake a study on the use of Artificial Intelligence (AI) in anticipatory actions as it relates to improving the health and wellbeing outcomes for children affected or at risk of climate impacts in Kenya.

The overall purpose driving this search for a consultant is to guide Barnfonden, ChildFund and SVEP in identifying the problem/need, and in finding a sustainable, appropriate, feasible and empowering AI solution to ensure children, their families and the wider environment are prepared for and know how best to address anticipated climate-related disasters.

The objective of the assignment is to investigate the use of, or potential use of, Artificial Intelligence (AI) and how it can be used locally in anticipatory action, specifically as it relates to improving the situation for children affected or at risk of climate impacts, by conducting a field study in either Samburu or Marsabit in Kenya related to their immediate and recent experience with a disaster occurring.

Background on Need for Study: 

Why Anticipatory Action and a Focus on Children: 

According to FAO’s Anticipatory Action and Response Plan (August–December 2023) published on 7th August 2023, the question is no longer if El Niño will happen, but what we must do to mitigate its impacts. Kenya is one of the 25 priority countries considered to be at high-risk of El Niño impact by FAO. This prediction indicated a greater than 90 percent chance of El Niño happening from October 2023 and continuing through to the end of the year. In Kenya the prediction has indeed come true as the country is experiencing heavy flooding due to the El Nino effects which has led to flooding in 23 out of the 47 counties

Natural disasters such as the ongoing El-Nino rains in Kenya, present a significant and growing threat to the well-being of children. When floods, drought, fires, cyclones and heatwaves hit, children not only suffer from hunger, but also other household deprivations. They drop out of school, are more susceptible to disease, experience higher levels of violence, and engage in more aggressive behaviour. According to a 2023 UNICEF report, children in 98% of African countries are at high or extremely high risk of the impacts of climate change. In Kenya, more than 16 million children – 67% of the country’s children - are living with the dual impacts of poverty and the climate emergency, according to new research by Save the Children. Drawing on data from the Children’s Climate Risk Index (CCRI) developed by UNICEF, the report emphasizes how children are biologically less equipped to handle the effects of shocks such as water scarcity to vector-borne diseases and flooding, to name just a few. The frequency and intensity of these shocks is increasing because of climate change. The report also revealed a major shortcoming in the extent to which international climate finance is responsive to children's needs. Just 2.4% of Multilateral Climate Funds (MCFs) funding was found to be allocated to projects that are child responsive. This assignment will hence focus on children as they are the group that faces extraordinarily high levels of exposure and vulnerability to climate-related hazards. Secondly a focus on children will help reduce a homogenous response by government and humanitarian actors when climatic disasters are predicted to occur. However, a system is required that can ensure children are involved (participate) in planning, are informed, empowered and are part of the solutions. ¨

Why AI: 

Preparing for and addressing these challenges require an urgent collaborative effort, informed and enabled by the latest science and technological solutions. Drought/floods early warning systems exist, but the crises provoked by climate change is often simply boxed as a food issue and not seen as a multidimensional crisis. New and emerging technologies such as AI can support government and humanitarian efforts before, during and after climate disasters occur. The use of innovations in tech such as the use of AI can help us move from reaction to anticipation by enabling earlier, faster and potentially more effective humanitarian action. Artificial intelligence can facilitate analysis and interpretation of vast and complex humanitarian datasets to improve projections and decision-making. An AI based technology can lead to better access to information, assistance and livelihood support, and facilitate stronger, more relevant needs analysis, a more prioritized and people/child-centred response, and more meaningful and systematic monitoring. Anticipatory action enables organizations to get ahead of a shock and mitigate its impact on vulnerable people. Predictive analytics can help anticipate community needs arising from different shocks. This opens a window of opportunity allowing humanitarians and government to reduce the overall impact of shocks by acting before needs materialise. Emerging evidence shows that anticipatory action is a more dignified, rapid, and (cost-) effective humanitarian response. This leads to the main research question posed by this assignment: How can AI be used to inform, curate, closely monitor and/or respond to effective, child-focused, anticipatory action strategies responding to potential and ongoing climate threats i.e floods/droughts, and encourages children’s participation and the engagement of duty bearers to inform, prepare and engage?

Anecdotal feedback from local partners is that AI could play a significant role in anticipating not only the occurrence of an event (for which some apps already exist), but in improving the preparedness and coordination of an anticipatory action response that more directly targets children and their families. AI might also be used to present information and provide advice related to the external factors impacting children and families in the lead up to and during disasters (things they may not directly influence), or the intrinsic factors – for instance, family and child psychosocial health or capacity to cope.

Working at community level, ChildFund and Barnfonden’s local partners interact on a daily basis with children, their families and those who support their livelihoods, health, education and protection: the agricultural extension officers, health centre volunteers, child protection services and school principals. Their plea is this: “We need a system that alerts us before climate disasters strike and helps us come together to create a more coordinated system of identifying impending disaster as it might affect children. This will allow us to pre-empt the worst effects of calamitous events so we can work together to keep children healthy, educated and safe.” They see the need to break down the silos, and to act with more speed ahead of risks – in the terminology of the UNHCR, to take Anticipatory Action[1]. One reason this makes sense is the calculation that every dollar invested in anticipatory action could give families seven dollars in benefits and avoided losses. This statistic alone has given us an impetus to find a cost effective approach that utilises advances in information and technology to develop solutions that will benefit children and their caregivers. Our local partners appreciate the efforts of Disaster Risk Management Authorities but want ways of prompting them into action earlier, or ways in which children and their families can be, themesleves, more directly engaged and responsible.

ChildFund Kenya’s child-focused response has been largely concentrated around schools and early childhood development centres, and on providing support to the most vulnerable families. Their El-Nino response interventions have focussed on the targeted areas where ChildFund Kenya has operations which include, Kisumu County (Kano plains Ahero and Nyakach Area), Migori County, Busia County, Nairobi County (along Nairobi River), Murang’a Landslides, Othaya Landslides, Meru Landslides, Isiolo county (Ewaso Nyiro River) and Samburu County others counties at risk include Marsabit, Turkana, Baringo and parts of Kajiado (ChildFund Kenya Flooding Sitrep, 2023). The Consultant will be guided by ChildFund Kenya on which of these areas the study will be undertaken. The Consultant will conduct a field-based study on information problems and gaps in Anticipatory Action, specifically as humanitarian responses apply to children, and how AI might be used to improve anticipatory action and/or its outcomes. This study will be conducted within a month of an El Nino event having occurred to ensure the event is fresh in people’s minds. The current El-Nino in Kenya provides a good standpoint to inform the recommendation of an AI solution that can be deployed before a climatic disaster to protect children`s wellbeing and health.

Purpose and Objectives of the Assignment: 

The objective of the assignment is to investigate the use of, or potential use of, Artificial Intelligence (AI) and how it can be used locally in anticipatory action, specifically as it relates to improving the situation for children affected or at risk of climate impacts. A field study will be conducted in one community in Kenya related to their immediate and recent experience with a disaster occurring. The ongoing El- Nino will be taken as an example of the disaster to be examined with the hopes that recommendations provided can be applicable to other types of climatic risks and disasters as it affects children wellbeing.

Key Study Questions:

Related to Anticipatory Action being targeted to children, are there gaps in the current system in relation to children’s safety and security that could be improved by AI-initiated data – what data would be considered advantageous; who should be receiving this data; how should they be receiving it; and to what end?

The study should answer the following key questions;

1. What specific gaps exists in the current Anticipatory Action system that targets children at risk of climate related disasters, and how can AI-initiated data contribute to addressing these gaps effectively?
2. Which types of data, if initiated by AI, would have been most advantageous as an anticipatory action for/with children? Please provide concrete examples and scenarios that have led to your recommendation.
3. Who are the primary stakeholders and end-users that should receive AI-initiated data for anticipatory action focused on children? Identify key entities such as the children themselves, their parents, local government, schools, health authorities, NGOs, and communities.
4. How might the recommendations, above, fit within the current humanitarian action system present at community level?
5. In what ways and through which channels should AI-initiated data be delivered to ensure maximum impact of anticipatory action for children? Consider factors such as accessibility, timeliness, and comprehensibility.
6. How can AI-initiated data be strategically utilized to improve anticipatory action outcomes for children? Explore potential applications in decision-making, resource allocation, and response coordination.

The study is aimed at generating data and providing practical recommendations that can be used by ChildFund Kenya, Barnfonden and their affiliates to strengthen the climate and disaster resilience of communities, specifically for children in vulnerable regions in Kenya that are affected or are at risk of climate disasters.

Scope of Work: 

The Consultant should be further guided by the following questions: 

1. Identify apps/information/data sources utilized by authorities (e.g. local government, schools, health authorities, NGOS) to inform their intervention logic and Anticipatory Action response in [this community]?
   • In what way were these sources helpful?
   • What critical elements or information were missing from the existing sources that, if available, would have made a response more effective or targeted?
2. How would AI be effectively integrated/used into the existing context to enhance anticipatory action response?
3. What system is currently in use to coordinate and activate community-level emergency response, in particular actions targeted at, and/or to, children and families? Highlight key components, strengths, and potential areas for improvement in the existing system.
4. Given the findings on Questions 1-5, what could you recommend as an AI solution (tool) to address current identified information/need gaps? Provide a detailed rationale for each recommendation, linking it directly to the observed shortcomings.
5. What data would you need to be able to confidently provide information to the proposed tool (information/need gaps; local, national, international etc)? Please also consider local knowledge / data (including weather data and indigenous/existing ways of knowing, consultations with agricultural extension workers, health volunteers and school principals, for instance, and how might we get it)
6. How can this tool be effectively utilised as a future response strategy to best support children, youth and their families (putting power in the hands of the people), as well as authorities working with them, to plan and respond?
7. What institutional support might be available to ensure the sustainability of potential AI app use? Consider factors such as training, maintenance, updates, and collaborations with relevant stakeholders.

Deliverables: 

Expected tasks of the Consultant shall include;

1. Pre-liminary meetings with key stakeholders (ChildFund Kenya, Barnfonden)
2. Conducting field visits and meetings with different respondents
3. Prepare and present a draft report with key findings
4. Drafting a comprehensive report of 20 to 35 pages

Timeframe: The exercise is expected to take approximately 4-5 weeks; commencing on 12th December 2023.

Management and Reporting Channel: 

Under close guidance of ChildFund Kenya`s team, the Consultant will directly report to the Programs and Sponsorship Director. The Consultant will work close interactions and consultation with ChildFund Monitoring, Evaluation, Learning and Knowledge Management Manger, Senior DRR Specialist and Barnfonden MEL Advisor and Programme Director.

ChildFund Kenya will provide the following:

1. Provide relevant project background information to support the assignment,
2. Facilitate mobilization of required respondents, local administration and leaders as may be necessary,
3. Provide day to day coordination of activities and as need arise,
4. Review the work done and provide feedback in a timely manner and
5. Pay the agreed costs for the assignments and as per the deliverables completed and contract term.

Terms of Payment: 

The Consultant will be paid as per the schedule below and terms stipulated in the contract for this assignment.

• Delivery of inception report 50%
• Delivery and approval of final report 50%

1. Compensation shall be paid NET, within 30 days from receipt of a proper invoice unless otherwise specified.
2. Payment will be made by cheque unless otherwise specified.
3. The payment shall be subjected to 10% withholding tax as required by the Law at the time of payment.
4. The full costs include professional fees for entire assignment.

Required Qualifications and Experience: 

• Experience evaluating/researching disaster-related anticipatory action or early action/early warning interventions strongly preferred
• Demonstrated experience in qualitative research methods and analysis required
• Experience leading research in humanitarian/disaster relief programs in Samburu or Marsabit counties strongly preferred
• Demonstrated experience conducting interviews/collecting data with diverse stakeholder groups required
• Experience in conducting research using information technology solutions, digital platforms, web based and mobile applications, innovative technology designs or use of Artificial Intelligence
• Fluency (written and spoken) in English & Swahili. Knowledge of the context of the project intervention zones highly preferred;
• Knowledge of disaster management national stakeholders at the County level;
• Strong ability to synthesize learnings in a clear and participatory way strongly preferred Excellent written communication skills with ability to deliver a cohesive and well structured report and visual presentation required

Application Materials: 

Candidates should include the following documents/information to demonstrate their qualifications

1. Technical proposal for expression of interest detailing the Consultant ’s interpretation of the task as outlined in the Terms of Reference
2. A brief financial proposal and work plan detailing the activities and number of days proposed for this assignment,
3. Curriculum vitae (CVs) of all team members who will be involved in this assignment (if institution or team of individuals),
4. Reference contact information of at least 2 individuals or institutions in the past similar assignments.
5. Examples of previous evaluation/research reports or research products in relation to the subject

[1] Anticipatory action is commonly defined as acting ahead of predicted hazards to prevent or reduce acute humanitarian impacts before they fully unfold (See Anticipatory action | OCHA (unocha.org)). UNOCHA, Governments and NGOs have agreed to increase their response to predicted disasters by taking Anticipatory Action as a means to reduce the impact of climate change and extreme weather events. This is the case in Kenya, where county governments have developed Anticipatory Action Plans (see for instance Stakeholders develop El-Nino preparedness plan – Kenya News Agency) in response to the predicted El Nino 2023 weather patterns and in support of the Kenyan government’s plans (see ChildFund Kenya’s El Nino Anticipatory Response Plan, Annex 1).

How to Apply: 

Interested Consultant s and firms should send their technical and financial proposals to kenyaprocurement@childfund.org by 8th December 2023

ANNEX 1: ChildFund International Humanitarian Situation Report

Humanitarian Situation Report

[Floods Emergency Response – Kenya]

[October 2023]

Part 1: The Overall Situation

The Kenya Meteorological Department forecasts a high probability (>80 per cent) of above-average rainfall from October to December in the whole country driven by warmer-than-average Sea Surface Temperatures (SSTs) over the central and eastern equatorial Pacific Ocean, indicating the presence of El Niño conditions. The situation may be exacerbated by a shift to a forecasted positive Indian Ocean Dipole (IOD), which typically enhances wetter conditions. The forecast shows that severe floods is expected in Kisumu, Migori, Bungoma, Busia, Nairobi, Mombasa, Lamu, Kilifi, Taita Taveta, Kwale, Tana River, Garissa, Wajir, Mandera, Turkana, Murang’a, Nyeri, Marsabit, Samburu and Isiolo. Areas likely to experience landslides include West Pokot, Kericho, Elgeyo Marakwet, Mt. Elgon, Narok, Nakuru, Baringo, Muranga and Makueni. An increase in water and vector-borne diseases including cholera, malaria, and Rift Valley Fever (RVF).

Based on this forecast, the Kenya Metrological department announced on 30th July 2023 that Kenya will experience El-Nino phenomenon, with resultant flooding from October to December 2023. However, on 23rd October President Willian Ruto announced that the country may not experience El Nino rains as earlier predicted but rather significant rains whose impact may not be as devasting as earlier projected. Despite the revised government on the likelihood of El-Nino, previous experience has shown that Kenya is in one of the regions where flooding has historically been higher when the October, November and December (OND) is predicted to have above average rainfall. During such previous above average rainfall or El Niño episode (like in 2019), severe flooding and massive landslides led to the destruction of property and essential infrastructure, crop and livestock losses, and increased epidemics, particularly of cholera. The 2019 El Nino for example affected more than 330,000 people in the country and resulting in the displacement of 160,000 people. As Kenya is emerging from a drought emergency, described as the worst in 40 years, the forecasted El Nino will prolong the suffering of the communities already struggling to recover from drought impact. The El Niño's destructive effects represent a major threat to access to food, drinkable water and other basic needs. The Arid and Semi-Arid Lands (ASAL) counties are at particular risk, with 2.8 million people suffering from acute food insecurity (IPC Phase 3 or above) following back-to-back climatic shocks. Wetter than usual conditions are also likely to result in epidemics such Cholera and malaria cases., Although the risk to communities living in flood-prone areas of the ASALs is particularly severe, the increase in rainfall corresponds to the usual short rainy season and could support recovery of the prolonged drought season and could improve food and nutrition security.

In the last two weeks, heavy rainfall have been experienced in Marsabit and Samburu Counties which led to flooding, destruction of several homes, vehicles being washed away. A number of deaths were reported in Karare and LogLogo wards of Marsabit County and Elbarta, Ndoto and Wamba West wards in Samburu County. On 28th October a vehicle for the Marsabit County Department of Agriculture was swept away by flash floods between Kargi and Mt Kulal in Laisamis sub-county following heavy rains currently being experienced in the county with the challenging road conditions. Thus, with the ongoing OND increasing in intensity in a number of counties across the country, the need for early action has become even more critical. The early action to be effective, efforts must be concentrated on preventing damage and loss to crops, livestock, productive lands, waters, and infrastructure to protect food at its source. This will not only safeguard local food supply but also mitigates wider effects on communities, local economies, and humanitarian aid requirements. FAO studies show that every USD 1 invested in anticipatory action can create a return for farming families of more than USD 7 in avoided losses and added benefits.

Identified Sectors for interventions by the Humanitarian partners (National and County Governments and Non-state actors)

Each sector has outlined responses in terms of the lifesaving interventions as well as intermediate recovery / livelihood activities. Preparation activities to mitigate the effects of the enhanced rains started in the month of September whereas response to the hazards was expected from October and towards November and December. Other medium-term interventions to address the aftermath of the enhanced rains will be implemented as from November and December, instead of October, as earlier, and stretch into January January 2024. Community preparedness will be enhanced through communication of early warning messages through National Government authorities, Sub-County level coordination forums, and other partners operating in the disaster affected regions. The key sectors of focus are Agriculture and livestock, health and nutrition, education, shelter and non-food items, food security, Water, sanitation, and hygiene.

Ongoing interventions by Humanitarian Partners: The humanitarian Partners are preparing plans to complement the Government of Kenya's response. There has been strengthening of the coordination mechanisms including activation of 8 Humanitarian Hubs. Early Warning Systems including advisories and messaging are being issued in collaboration with government sectors. Support for community disease surveillance particularly in urban informal settlements. There is the support for training on the Kenya Inter-Agency Rapid Assessment (KIRA) and plans to carry out assessment in hot spot areas. Most partners are also prepositioning essential non-food items in strategic Hubs.

Part 2: Areas Where ChildFund Works that anticipate flooding

The areas expected to be impacted by significant rains where ChildFund operates are; severe flooding is expected in Kisumu, Migori, Busia, Nairobi, Turkana, Murang’a, Nyeri, Marsabit, Samburu and Isiolo. Floods have already been reported in Marsabit and Samburu Counties. Areas likely to experience landslides include Elgeyo Marakwet, Narok, Nakuru, Baringo, Muranga and Makueni. An increase in water and vector-borne diseases including cholera, malaria, and Rift Valley Fever (RVF) is expected to be witnessed in various counties; Kisumu, Migori, Busia, Nairobi, Turkana, Marsabit, Samburu and Isiolo. is anticipated that the enrolled and sponsored families will be affected.

Part 3: ChildFund Kenya’s Response Preparedness

ChildFund Kenya has put in place an Emergency Preparedness Plan (EPP) and flood contingency plan in place to inform preparedness and response. Various Local Implementing Partners (LIPs), including Nawiri Child Development Program and Samburu Children Program, Western Community Children Program, Nairobi Metropolitan Program and Central Rift Children Development Program have since developed their own contingency plans with two partners having their EPP updated as an anticipatory measure to respond to the floods. All the Local Implementing Partners (LIPs) have been capacity built on Disaster preparedness and contingency planning. ChildFund is part of the coordination forums and technical working groups to plan for El Nino response including Kenya Humanitarian forum, WASH sector working group, Cash Working group Education and Emergency (EiE) and Food security cluster while the partners participate in the county forums including County steering group meetings and county sector working groups. ChildFund in partnership with WeWorld carried out a rapid assessment in Isiolo county to document the post drought gaps and preparedness actions for floods in Isiolo. The previous SitRep for the Month of September, ChildFund Sweden (Barnfonden) has allocated USD 33,000 to ChildFund Kenya for anticipatory action.

The expected preparedness and Response Sectoral Needs and Gaps

Through the Kenya Humanitarian forum and government of Kenya contingency plan, various preparedness and response gaps have been identified as listed below.

Food Security and Livelihoods

• Expanded school feeding programme to reach all children affected by the various emergencies.
• Food distribution to affected households.
• Cash transfers for the affected households. Provision of cash transfers to support households in acquiring their most urgent needs to enable them to cope with the situation.
• Provision of subsidized planting and top-dressing fertilizer to households in affected areas.
• Provision of certified seeds: Provision of Local/indigenous and high value crops

Livestock sector

• RVF awareness creation training and planning
• Control and surveillance of emerging livestock diseases: Rift Valley Fever (RVF) and others.
• Procurement and administration of RVF Vaccines
• Restocking / redistribution (Livestock market)

Health and Nutrition Sector

• Health promotion - awareness about waterborne diseases and how to keep safe
• Provision of treated mosquito nets to mothers and children under five years
• Indoor residual spraying will be conducted in malaria epidemic counties.

Education

• Rapid assessment on children affected by the floods (disaggregated by gender); and education facilities and learning materials
• Provision of makeshift schools (tents) and learning materials
• Support children with scholastic materials
• To rehabilitate collapsed school infrastructure to minimise disruption of learning.
• Provision of teaching/learning materials including education kits and recreational kits.
• Advocacy: ensure that child protection concerns are addressed in flood areas.

WASH Sector

• Provision of water treatment chemicals and testing kits
• Undertake early recovery and rehabilitation of destroyed WASH infrastructure. Provision of storage tanks (collapsible and plastic) at strategic positions.
• Procure spare parts, service, and repair damaged gensets. Relocation of water supply points
• Sanitation and hygiene promotion and proper disposal of dead carcasses
• Rehabilitation of dams, pans, and shallow wells

Shelter and Non-Food Item sector

• Family Kits and dignity kits to take care of the affected households.
• Mosquito nets, blankets, malaria drugs, nutrition supplies for the vulnerable groups, household water treatment due to fecal contamination.
• Sheltering the flood victims temporarily
• Need for relocation to safer places
• Need to establish alternative routes
• Construction of houses after the floods

Security and Logistics

The security situation across the county has generally remained calm however, in the areas expected to experience significant rains, isolated incidences of resource based intercommunal conflicts have been reported especially in the ASALs, however it is anticipated that movement may be disrupted during the rainy seasons as roads are cut off.

Partnership and coordination

ChildFund and its LIPs continue to participate in national and county level emergency response coordination forums. LIPs are actively participating in county level steering groups charged with assessment, planning and implementation of interventions. ChildFund also works in collaboration with other international and local Non-Governmental Organizations and UN agencies such as UNICEF FAO, Save the Children, Mercy Corps, USAID, World Vision, WeWorld etc.

Staffing

Ongoing emergency Preparedness and response in the target areas is currently supported by Country Office and LIP staff. One Emergency Response Officer is currently available with also various emergency focal staff at the LIP level.

Donors:

In the face of various disasters in the various counties no major funded response project has been undertaken, however ChildFund Kenya continue to get support from ChildFund Alliance members and international office to support the affected families.

Media/Communications

Plans for collecting photos/videos/stories, i.e. should an outside photographer be hired?

Normally when implementing such emergency grants, the officers in the field are provided with HD cameras for collecting quality photos and videos and providing success stories from beneficiaries who have benefitted and become resilient out of the intervention. The donor at times can arrange for the outside media services to collect videos and stories in the form of documentary meant for visibility and fund raising.

Key points for messaging and visibility, particularly any host government sensibilities that must be considered.

Being guided by humanitarian principles and measures, the information for visibility either by print or broadcast media must meet the threshold of the standards that govern it either internationally or locally.

Support needed or requested from IO, GSS or Global Teams – whether onsite or remote.