All projects
Case study

CyStellarSatellite Risk Dashboards, Bilbao Fire Service & Roof AI

Dashboards for CyStellar's satellite risk platform, an emergency forecasting dashboard for Bilbao's fire service, and AI that maps roof materials across Europe.

Role
Frontend Developer Intern
Timeline
Jul 2022 - Dec 2022
Type
Project I worked on
Stack
React, JavaScript +6
cystellar.com
01

Overview

CyStellar is a space-tech company, based in London with an office in Sweden, whose platform, TerraRisk, turns satellite Earth observation, weather forecasts and hazard data into risk insights for insurers, and helps public agencies detect and monitor risks. I joined as a front-end intern from July to December 2022 and worked on three things. The first was CyStellar's internal dashboard, putting geospatial and environmental data on interactive maps and charts. The second was CyStellar's project for the City of Bilbao in Spain, run through REACH, an EU Horizon 2020 incubator. Bilbao City Council, which runs the city's fire brigade, wanted to size and deploy its firefighter teams better, using its history of mobilisations, interventions and emergencies alongside weather alerts, city events and points of interest. CyStellar's answer was a cloud geospatial platform that forecasts whether each type of emergency will rise or fall over the next 15 days, with a dashboard for the fire service's managers. The third was machine learning: training models to classify the roof materials of houses across Europe from satellite imagery. CyStellar's internal tools aren't public, so the platform images here are CyStellar's own, alongside photos of Bilbao.

02

Highlights

  1. 01Front-end work on CyStellar's internal dashboard for TerraRisk, its satellite risk platform
  2. 02Interactive maps and charts with Mapbox GL and D3.js, layering risk and environmental data over satellite imagery
  3. 03Work on CyStellar's project for Bilbao City Council, through the EU's REACH incubator, to help size and deploy the city's firefighter teams
  4. 04Forecasts of whether each type of emergency will rise or fall over the next 15 days, built on Bilbao's history of firefighter mobilisations, interventions and emergencies
  5. 05Weather alerts, city events and nearby points of interest as context for those forecasts
  6. 06A decision-making dashboard for the managers of Bilbao's fire service
  7. 07Machine learning models trained to classify the roof materials of houses across Europe from satellite imagery

Next project

Dokan.gg | Multi-Vendor Marketplace

A multi-vendor e-commerce marketplace for Bangladesh, with a storefront for shoppers and dashboards for sellers and admins.

View case study