HydraMind

Kolo Naukowe Neuroinformatyki "Neuron"

  • 0 Raised
  • 54 Views
  • 0 Judges

Tags

  • Poland

Categories

  • Challenge #3: Disaster risk monitoring​

Gallery

Description

1. Idea / Project Summary

  • The Idea: To create a live, queryable intelligence platform that moves beyond static catchment maps by fusing orbital data with subterranean municipal infrastructure.

  • The Problem: MPWiK S.A. currently operates in a reactive mode. They lack objective, city-scale data on land cover and permeability, which is essential for calculating surface runoff, designing nature-based solutions, and calibrating the hydrodynamic catchment models used for decision support.

  • The Solution: A Dual-Layered Intelligence Pipeline that integrates:

    • Stream A (Orbital): Automated queries for cloud-free multispectral imagery to generate objective land classification.

    • Stream B (Subterranean): Direct ingestion of MPWiK’s geodata (3,314km of sewers and ditches) to map mathematical connectivity.

    • Actionable Output: A high-performance MapLibre GL JS Dashboard that identifies "GBI Opportunity Hexagons" - specific 300m zones primed for nature-based solutions like swales and rain gardens.

2. EU Space Technologies

The platform leverages the Copernicus ecosystem to provide objective environmental context:

  • Space Data: Uses Copernicus Sentinel-2 L2A multispectral imagery.

  • APIs & Access: Automated data retrieval via the Copernicus Data Space Ecosystem (CDSE) STAC API.

  • Spectral Analytics: Processes raw bands (B02, B03, B04, B08, B11) to calculate indices including NDVI (vegetation), NDWI (water), and MNDWI (modified water index).

  • Processing Frameworks: Built on a modular Python Analysis Stack using geopandas for spatial logic, shapely for geometry processing, and networkx for topological network diagnostics.

3. EU Space for Water Challenge

This project directly addresses Challenge 3: Using satellite data for land cover classification and green-blue infrastructure design and planning.

  • Land Cover Assessment: It provides an objective estimation of surface runoff by classifying 4.5k grid cells into functional tiers, ranging from permeable green space to high-impermeability industrial zones.

  • Water Management Support: The platform automates "Topological Network Diagnostics," identifying over 21,000 isolated segment anomalies and queuing them for municipal review to improve system resilience.

  • GBI Planning: By calculating the "GBI Gap Equation" - factoring in drainage density, service zones, and existing swales - the tool identifies exactly where nature-based solutions will have the highest impact on stormwater retention.

4. Team: HydraMind

Denys Kovzikov

Camille Lisek

Tymek Marzec

Pranaya Anargya

Daniel Vasquez

Anna Nowak

5. Prototype / Demo

  • Live Experience: The dashboard features a MapLibre GL JS environment rendering geometries without lag. It includes interactive sidebar controls for the Brochówka (26.4 km2) and Ługowina (13.9 km2) catchments, alongside live analytics for pipe diameter distribution and land cover breakdowns.

GitHub Repository link: https://github.com/KN-Neuron/Cassini-HydraMind-dashboard.git

Attachments