Over the last several weeks, we have witnessed massive flooding from Texas to Pakistan to the DRC. Hundreds of lives lost and billions of dollars in property damage.
This prompts the question: how can we rapidly and effectively build the infrastructure needed to warn communities of imminent crisis - in a way that’s inclusive, scalable and sustainable? A recent roundtable hosted by British Expertise International (BEI), People in Need (PIN) and other thought leaders explored this question and took a deep dive into what’s working across Southeast Asia where localised responses, private sector engagement and digital innovation are reshaping disaster risk reduction strategies.
PIN helped set the context for the discussion, sharing the remarkable journey of Cambodia’s early warning system (EWS) from its humble beginnings to a sophisticated, nationally integrated platform.
National Systems, Local Realities
The evolution of Cambodia’s early warning system (EWS), began as a community-based, manual effort: villagers raising flags to signal danger. PIN described how EWS has since grown into a nationally integrated, sensor-driven platform used by the government. This transformation was made possible through a combination of locally-appropriate technologies such as IVR (interactive voice response), locally made and low-cost real time data sensors and integration with Cambodia’s social registry (ID Poor 1 and 2) to ensure that the most vulnerable receive timely alerts. In Laos, a more top-down model has emerged, while Nepal is now working to decentralise its national system to improve responsiveness at the provincial level by using impact-based forecasts as well as multiple dissemination channels tailored for a specific area. The conclusion is clear: functional national systems must be rooted in local realities. That means designing tools that are low-cost, easy to maintain, and accessible across digital divides. Inclusivity – in both access and design – is not an afterthought but a core feature of any system that aims to save lives.
To bring these initiatives to scale, participants largely agreed that we also need donor buy-in, in particular flexible, long-term funding. Donors including FCDO are leading the way, recognising the growing threats posed by multi-risk hazards as climate change intensifies. For example, FCDO has committed to a five-year investment in PIN’s RAIN programme, which aims to support early action and strengthen climate resilience for vulnerable and marginalized communities facing multiple hazards in Nepal’s Madhesh and Lumbini provinces by 2029.
As was also highlighted during the roundtable, initiatives such as the Cyclone Preparedness Programme in Bangladesh and Somalia offer valuable lessons on integrating community preparedness with formal national systems, including shock-responsive social protection.
Moving from nationally-owned to locally-led early warning and adaptation
Partnerships have been central to the scale-up of early warning systems in Cambodia, Nepal and beyond. PIN described how they work closely with governments, mobile operators and private sector actors, including Google, Smart Axiata (a telecommunication company in Cambodia), and eSewa (a bank in Nepal), to co-develop systems that are low-cost, interoperable and built on open-source technology. In Cambodia, alerts are disseminated not only through SMS and IVR but also via telecom providers; there is also an experiment toto send alerts through banks, which provide an emerging channel with vast reach. In Nepal, conversations with eSewa, a bank with 10 million users, are opening doors to new dissemination pathways. One of the clearest lessons from this experience is that effective partnerships must start early, with shared incentives and a mutual understanding of roles and not be an afterthought after systems have already been designed.
Social inclusion in early warning and response systems
Digital solutions alone are not enough, especially in crises where trust, access and comprehension vary widely. In the Rohingya camps, for example, mobile connectivity is restricted, so radio, face-to-face communication and trusted community figures become critical to effective early warning. Similar approaches are being explored in Malawi and Mozambique, where local leaders remain central to disseminating warnings in low-connectivity environments.
Inclusion can be integrated from the outset through multilingual alerts, multiple modalities as well as accessibility features for people with disabilities. Trust is not built through a single alert; it is earned through repeated and consistent lived experience. Understanding how communities share and validate information internally is just as important as the technology. A truly inclusive system is one that works for everyone, not just those online or within network coverage.
DRM strategies can be designed not only to prevent losses but actively to strengthen local economic resilience particularly in fragile or informal economies where livelihoods are most vulnerable to climate-related shocks. The more local or national market actors are engaged in DRM activities, the greater the contribution to strengthening economic resilience.
Enabling Growth
Data from early warning systems can inform hazard and risk mapping, an essential foundation for building climate resilience.
To ensure resilient economic growth, development finance must be guided by a clear understanding of climate vulnerability and risk. This includes assessing the potential impacts of climate change on local market systems and ensuring that investments strengthen or diversify value chains for those most exposed to climate-related shocks.
Early warning systems provide critical data that can support not only communities and governments in identifying vulnerability, but also private sector actors including insurers and financial institutions. For these stakeholders, early warning data can inform actuarial models, reduce uncertainty in underwriting and guide lending or investment decisions in high-risk areas. By integrating early warning data into financial planning, the private sector can better manage exposure and promote more adaptive, resilient economies.
AI - What Comes Next?
Artificial intelligence has the potential to improve simulation, the accuracy of forecasting and scenario-based planning, particularly when embedded into human-centered design.
Participants highlighted that several innovation labs, including those supported by development finance institutions, low-cost sensor development and AI-driven flood forecasting tools are being piloted. Promising applications include AI-generated sign language alerts, hyperlocal forecasting in partnership with Google and simulation tools to test and refine emergency response. But AI is not a silver bullet. It should strengthen rather than replace human validation, trust or community feedback mechanisms. The value of AI lies in its ability to augment planning and scale decision-making, not to automate away the relationships on which effective systems depend.
Who is responsible for warning dissemination and response?
Early warning systems are no longer just a technical fix: they have become a human right. As recent events have shown, they save lives, protect homes and safeguard communities. The path to a climate-resilient world runs through communities themselves. What we build must be locally led, digitally enabled and socially trusted.
Participants shared that sustained donor investment remains a key enabler, with examples from FCDO’s multi-year support illustrating the value of long-term funding in scaling community-rooted systems.
%20(2).jpg)
