The project aimed to investigate how the two emerging technologies - Internet of Things (IoT) and Big Data technologies can be used to provide a digital platform for monitoring both flooding and drought disasters. Further, it also ventured to explore sensor-based collection of data to allow development of predictive algorithms for disaster prediction and management. Although, our initial studies examined natural disasters encompassing geophysical, meteorological, hydrological, climatological and biological calamities, our pilot focus concentrated on flooding and drought. Flooding is one of the major hydrological disasters that have perennially affected most parts of Kenya during the rainy season. Drought on the other hand has ravaged many of the arid and semi-arid areas of the country and sometimes, due to climate change, threatened the food security of the previously unaffected areas.

A common characteristic of both of these disasters, is the difficulty in obtaining timely and relevant information with regard to the prediction of their occurrence, the identification of risks, disaster preparedness, emergency response, resource allocation and re-allocation, reaction planning and post-disaster reconstruction strategy. Furthermore, the participation of the public in providing information even in the form of pictures or related specific info that can contribute to the identification of an impending flooding or drought disaster is conspicuously missing. Technology, therefore, is seen to harbour a lot of potential that can provide a holistic platform that consists of both information from intelligent devices as well as the public user input. Hence, in our research quest, we studied the opportunity that IoT presents in collecting information such as the changing soil moisture, rain level, intensity of light, temperature, humidity as well as running water levels to help determine the patterns of flooding and drought.

As a technical pilot, we deployed a low power wide area network (LPWAN) known as LoRaWAN in the four partner institutions to be able to collect all these data at a low power network and a long range. Eight IoT nodes (two for each institutional site) packaged with sensors collecting the information were deployed and are currently running as we build our database for development of predictive analytics on the collected data.

(i) Upcoming

1. Further studies and grant sourcing for Machine Learning and Artificial Intelligence on the Data being collected from the sensors.

(ii) Previous

1. Analysis, design, prototyping and testing of an IoT-based big data ecosystem that can assists in management of flood disasters through defined metrics of flood indicators.
2. Implementation on a pilot basis, an IoT-based big data analytics ecosystem for disaster prediction and management: https://dpm-platform.web.app/ 
3. Testing and validation of the IoT-based big data analytics model for disaster prediction and management.

(i) Postgraduate Students

1. Mr. Leonard Mabele – Strathmore University (PhD Student and Lead Technical Team)-studies still on-going
2. Mr. Stephen Ng’etich – Strathmore University (MSc Student and Member Technical Team ) studies still on-going
3. Mr. Solomon Kamau – Strathmore University (MSc Student and member technical team)- studies still on-going
4. Mr. Kibet Alex – Laikipia University (MSc Student )- studies still on-going
5. Ms. Lisper Kendi – Meru University of Science and Technology (MSc. Comp science student & Research assistant) studies still on-going

(ii) Publications

1. Mugeni et.al., 2021: Determinants for the Adoption of Internet of Things (IoT) for Flood and Drought Disaster Management in Kenya (http://dx.doi.org/10.15863/TAS); ISJ Theoretical & Applied Science, 09 (89), 342-354. Soi: http://s-o-i.org/1.1/TAS-09-89-43 Doi: https://dx.doi.org/10.15863/TAS

(iii) Authors

1. Gilbert Barasa Mugeni, Ph.D ; Kelvin Kabeti Omieno, Ph.D ;Franklin Wabwoba, Ph.D; Simon Maina Karume, Ph.D; Leonard Mabele; Daniel Otanga, Ph. D.

Page last updated on 14.04.2022