Geohash-Based IDs: A Clever Pattern for Location Data

When building the NZ Speed Camera Map, I needed unique IDs for 896 cameras. The catch: no database, just coordinate data from a government CSV. Geohashing solved this elegantly.

The Problem

Government data includes coordinates but no stable IDs. Camera names change. Addresses vary. I needed IDs that:

• Stay consistent across data updates
• Require no central authority
• Allow proximity queries

The Solution

import ngeohash from 'ngeohash';
 
export function generateCameraId(lat: number, lng: number): string {
  return ngeohash.encode(lat, lng, 8);  // ~19m precision
}

An 8-character geohash like rbsm6p7q uniquely identifies a camera's location. Even if metadata changes, coordinates don't—so the ID remains stable.

Precision Levels

Practical Benefits

Deduplication: When scraping data, duplicates naturally collapse:

const uniqueCameras = Array.from(
  new Map(cameras.map((c) => [c.id, c])).values()
);

Proximity: Cameras with shared prefixes are nearby—useful for clustering.

SEO-friendly slugs: Combine geohash with readable info:

const slug = `auckland-cbd-${geohash.slice(0, 6)}`;

When to Use This Pattern

• Static datasets with coordinates
• No database available
• Need deterministic IDs from location data
• Want built-in proximity information

Code

The full implementation is in nz-speed-cameras on GitHub.