Masaar

After reading about Snoonu’s autonomous delivery initiative with QRDI, I got curious about the coverage problem underneath it. Where does drone delivery actually beat ground vehicles across greater Doha, and where doesn’t it? I spent a few days building a simulation to find out.

Highlights

Models two delivery modes from a depot you can drop anywhere across greater Doha’s real census zones, then colors each zone by whichever mode wins.
Drone ETAs use a 60 km/h cruise with a 5km range limit and a hard exclusion zone around Hamad International Airport. AGV ETAs apply a 1.3x road tortuosity factor over a 25 km/h average, with peak traffic penalties in commercial zones.
A greedy two-depot optimizer drops a second depot where it adds the most residual population coverage.
The takeaway: one central depot covers a surprisingly small share of the population by drone. The airport exclusion zone and 5km range bite harder than you’d expect, so AGVs win most zones by default and drones only pull ahead in the dense inner-city areas where straight-line distance is a real advantage.

Tech Stack

React + Vite - Frontend
Leaflet + OpenStreetMap - Map and CARTO dark basemap tiles
Qatar PSA census zones - Bundled district polygons and per-zone population
Turf.js - Geospatial math for distances and coverage
Tailwind - Styling

Notes

The zone boundaries are the official Qatar PSA census polygons (COD-AB admin level 2, via the Humanitarian Data Exchange), bundled with the app so the map is reliable with no live API calls. Each polygon joins 1:1 to the census by its official zone number.
Population weights come from real PSA census figures, joined per zone. The census runs a few years behind (up to 2019), which underrepresents fast-developing areas like Lusail and The Pearl, so I call that out in the methodology panel.
The name Masaar (مسار) means route or path in Arabic.
I built this before reaching out to Snoonu’s team. I wanted to understand the problem before the conversation.

Check it out: Masaar