AI-powered tools to manage water in Africa
Key Highlights
- Limpopo Digital Twin and WaterCopilot tools were rolled out to manage water across Botswana, Mozambique, South Africa, and Zimbabwe.
- The programs were part of a collaboration between the International Water Management Institute and Limpopo Watercourse Commission.
Limpopo Digital Twin and WaterCopilot digital tools were rolled out to manage water across Botswana, Mozambique, South Africa, and Zimbabwe following an International Water Management Institute (IWMI) and Limpopo Watercourse Commission (LIMCOM) collaboration to equip LIMCOM member states with artificial intelligence (AI) to manage climate variability and transboundary water challenges.
The Limpopo Digital Twin is an interactive digital replica of the river basin that delivers near real-time insights for water-planning. The first prototype Digital Twin in Africa, now fully mature and openly accessible, integrates hydrological, climatic, irrigation, and water-use data with intuitive maps and analytics.
WaterCopilot is an AI agent embedded into the Digital Twin platform. It was developed by IWMI in collaboration with Microsoft as a conversational interface to basin data. WaterCopilot allows water managers to query complex hydrological information using plain-language questions, like English or Portuguese, without needing specialist knowledge of the underlying datasets or platform architecture.
To support water managers’ use of the platforms, IWMI and LIMCOM hosted an April 15-16 capacity-building workshop in Mozambique which focused on testing and scenario development for these tools.
A live demonstration of the WaterCopilot showcased plain-language queries on reservoir status, environmental flow thresholds, water accounting, and MiniSASS river health scores which is a citizen science tool for monitoring water quality and health. Through the training stakeholders learned to navigate the Digital Twin, which provides near real-time operational monitoring with WaterCopilot making data accessible to users at all levels.
On the final day of the workshop, LIMCOM member states were introduced to the Water AI project, which is an AI-augmented global hydrological modelling initiative. Water AI can create simulations of what could happen under various conditions, while using AI calibrations to adjust the model to match real-world data as closely as possible using past observations like rainfall and river flow. The Limpopo River Basin will serve as the first real-world test of this AI-driven hydrological model.
During the workshop, technical representatives from LIMCOM member states co-developed priority scenarios for the Limpopo River Basin, designating it as the primary Water AI project demonstration case for Southern Africa.