Skip to main content

Design Specification

Document: EOHA Design Specification v2  |  Prepared by: Simamnkele Mlisana & Msesenyane Makhongela
Organisation: Medical Consortium of Africa (Pty) Ltd, University of Pretoria / TuksNovation

This section contains all design artefacts produced during Milestone 1. Each page renders the diagrams inline and provides download links for original files.

Document sections​

SectionContents
System Architecture →Pipeline architecture, data flow, component diagram
Data Architecture →ERDs — Observation Prediction and User Management
Use Cases →Use case diagrams, use case descriptions, actor roles
Wireframes →Home, Malaria, and NCD page wireframes
Operational Workflow →End-to-end operational workflow (UML activity diagram)

Download original files​

FileFormatLink
Design Specification Document v2PDFDownload →
Initial UML (Draw.io).drawioDownload →
Use Case SetWord .docxDownload →

System description​

The EOHA platform encompasses the full lifecycle from environmental satellite data ingestion through to actionable disease forecast delivery. It includes:

  • Data Ingestion — automated retrieval and mapping of EO satellite data and clinical ground-truth observations to spatial zones
  • Predictive ML — XGBoost algorithms producing Malaria_Risk_Forecast and NCD_Risk_Forecast scores per zone
  • Geo Dashboard — interactive Leaflet map with colour-coded risk zones, historical health data, and AI predictions
  • Alert System — monitors forecasts against predefined thresholds and dispatches targeted warnings to authorised health officials

Actors​

ActorPrimary responsibilities
End UserBrowse risk maps, view AI forecasts, track and acknowledge alerts
Data ContributorLog ground-truth observation data, report case counts per zone
AdministratorManage user roles, configure zone boundaries, oversee system access
System (automated)Ingest EO data, run ML predictions, render geospatial outputs, dispatch alerts