As heavy rains return to Kampala, an artificial intelligence-powered early warning system called Disaster Sense AI is being proposed to alert residents of flooding threats. The system uses ultrasonic sensors installed in drainage channels to monitor water levels and trigger SMS alerts when flooding threatens nearby communities, according to Eng. Dr Peter Byamugisha, executive director of Ardhi Advocacy and Research Innovation (ARDHI). Residents of flood-prone neighbourhoods including Bwaise, Nabulagala, Nalukolongo, Najjanankumbi and Kalerwe face recurring dangers when heavy rains overwhelm drainage systems.
Flooding in Kampala has caused significant harm. On October 11, 2026, heavy rains in the Kampala Metropolitan Area claimed two lives and damaged property including vehicles worth millions of shillings. In 2025, five deaths were reported in the metropolitan area, and the government pledged Shs20 billion in compensation for affected traders. The developers believe early warnings could give residents valuable time to move to safer locations, secure belongings and seek higher ground before water levels become dangerous.
Byamugisha explained how the system works: ultrasonic sensors detect rising water levels in drainage channels, and artificial intelligence then sends alerts to registered residents in affected areas, warning them to evacuate. However, the system's effectiveness will depend on reliable sensors, adequate coverage, accurate identification of threatened areas and residents' ability to receive and act on messages in time.
ARDHI has approached government authorities seeking approval to deploy the technology. Byamugisha said the organisation has written to the Office of the Prime Minister and other government departments to present the innovation for consideration. Kampala Minister Mincer Kabanda said the government is open to proposals that could help reduce flood damage. The developers also say the technology could be adapted to monitor landslide risks in vulnerable communities in the Elgon and Rwenzori regions.