Young innovators from drought‑prone regions of Kenya took the top two prizes at the ninth Young Scientists Kenya National Science and Technology Exhibition with projects that combine sensors, automation and artificial intelligence to tackle drought, water scarcity and food‑production challenges.
AI system forecasts drought risk
First Place Overall was awarded to Hafsa Ibrahim Adan and Asha Harun Yarow from Senior Chief Adano Girls Day Secondary School in Wajir County. Their project is an AI‑based environmental monitoring and drought risk prediction system designed for climate‑vulnerable communities in north‑eastern Kenya.
The system gathers data from a suite of environmental sensors — measuring soil moisture, temperature, humidity and rainfall — and feeds these inputs to an artificial intelligence model. The model produces a drought‑risk score, an estimated probability of drought, the model’s confidence level in that prediction and an estimate of how many days remain before drought conditions are likely to develop.
According to reporting on the exhibition, the intention is that the early warning information could give farmers and communities time to prepare irrigation systems, prioritise limited water resources and protect crops and livestock before conditions become severe.
Automated garden conserves water and protects crops
The Second Place Overall prize went to Halima Aila Mohamed, Sake Hussein Tadicha and Manal Mohamed of Zad Muslim School in Marsabit County. Their project, Eco‑Guard, is an automated smart garden aimed at conserving water, protecting plants and improving environmental cleanliness in an arid environment.
Eco‑Guard integrates a protected greenhouse structure with automated irrigation and temperature control. Key elements described at the exhibition include:
- a soil‑moisture sensor that triggers a water pump when the root zone becomes dry, and stops watering once the target moisture level is reached;
- a temperature sensor that opens an automated roof vent when internal heat rises beyond a set threshold;
- automation intended to reduce water waste and stabilise growing conditions for seedlings and crops.
Both projects respond directly to the realities of Kenya’s arid and semi‑arid counties, where timely information and efficient water use can make the difference between stable harvests and crop failure.
Why these solutions matter
Early warning systems and automated water management target two linked problems: the need to anticipate climatic stressors, and the need to stretch scarce water supplies while maintaining food production. Practical student projects such as these show how low‑cost sensors and machine learning can be combined to produce actionable information for households and smallholder farmers.
Advantages of the approaches presented at the exhibition include:
- localised monitoring that captures conditions at field or village level rather than relying only on coarse regional forecasts;
- automation that reduces the labour and attention needed to maintain irrigation and protect plants under heat stress;
- decision support — drought‑risk scores and timing estimates that help prioritise interventions before conditions deteriorate.
| Prize | School | County | Project |
|---|---|---|---|
| First | Senior Chief Adano Girls Day Secondary School | Wajir | AI environmental monitoring and drought risk prediction |
| Second | Zad Muslim School | Marsabit | Eco‑Guard automated smart garden |
These student projects are early‑stage, demonstrator‑type solutions that would need field testing, user training and maintenance plans before being scaled. Nonetheless, they illustrate how relatively simple sensor networks combined with machine learning and automated actuators can help communities manage climate risk.
At a time when many parts of eastern and southern Africa are grappling with drought, innovations from young scientists offer practical steps that can be trialled and adapted locally.
