CEOIS2 – Improvement, Adaptation and Implementation of previously developed Customised Earth Observation Services for Improving Water Supply Management in Uganda

The aim of this project is to improve, adapt and integrate newly developed services (CEOIS2) into Utility Performance Monitoring and Information System (UPMIS) which is daily used at GeoCodis's partner, the Ministry of Water and Environment (MWE) of the Republic of Uganda.
The main technical objectives are to improve, adopt and optimize existing build-up information services to Uganda conditions and to compare two methods of machine learning methodology. Final results will be monthly build-up images for close to 1000 settlements in Uganda and all needed REST-API that will be use within UPMIS application.
It is foreseen that satellite products will be generated by two different methodologies. Firstly, on a base of implementation of semantic segmentation architecture DeepLabV3 and secondly with the use of machine learning algorithm U-Net. Detailed specification which approach will be implemented and used within rural settlements and within different yearly season, will be tested and selected during the project.