Data Source Indicators
When the Planning Dashboard shows a building count, you may see a small icon next to the number. This tells you how the count was calculated — and why it matters for accuracy.
Why Two Methods?
The UK government publishes OS Open UPRN data — a list of every addressable property in the country with precise coordinates. AGMT uses a subset filtered to Leicester City loaded into a local DuckDB database in your browser.
This data lives in two files:
A compact file listing every unique street with its bounding box (geographic boundary) and total address count. Think of it as a table of contents — it tells you "High Street has 42 properties" without listing each one.
A much larger file containing every individual UPRN (property) with its exact latitude, longitude, and street name. This is the full phone book — every single address, one by one.
The two methods exist as a performance optimisation: if the tool can use the summary, it does — because it's instant. It only falls back to counting individual addresses when it has to.
The Two Methods Explained
Fast Path
Your drawn shape completely covers the entire street — from end to end. Since the whole street fits inside, AGMT grabs the pre-calculated total from the summary index. No need to count addresses individually.
Analogy: Imagine a bookshelf where each shelf is labelled "Shelf A — 42 books". If you need all of Shelf A, you just read the label instead of counting every book. That's Fast Path.
Instant
100% coverage
Uses summary index
No network request
Precision Scan
Your drawn shape only partially overlaps a street — maybe it covers one end but not the other. AGMT can't use the summary total (that would overcount), so it checks each individual address to see which ones fall inside your shape.
Analogy: You only need books from the left half of Shelf A. The label says 42 total, but you can't use that — you have to walk along and count only the ones within your section. That's Precision Scan.
How the Decision Is Made
When you draw a shape, AGMT compares it against every street's bounding box:
This decision is automatic — you don't choose a method. AGMT picks the fastest accurate path for each street individually. A single shape can have some streets via Fast Path and others via Precision Scan.
Both approaches are private — no data is sent to an external server. The processing happens on your own machine.
Side-by-Side Comparison
| Aspect | Fast Path | Precision Scan |
|---|---|---|
| When used | Entire street fits inside your shape | Only part of the street is inside your shape |
| Speed | Instant (single lookup) | Slightly slower (scans each address) |
| Accuracy | Exact — pre-calculated total | Exact — counted one by one |
| Data file used | Street Summary Index | Individual Address File |
| Coverage shown | 100% | Calculated percentage (e.g. 67%) |
| Network needed? | No — local DuckDB | No — local DuckDB |
| Region | Leicester City only | Leicester City only |
Coverage Percentage
When Precision Scan is used, AGMT also calculates a coverage percentage:
"Your shape covers 67% of this street (28 of 42 properties)"
This tells you how much of the street falls within your drawn boundary. Fast Path always shows 100% because the entire street is covered.
Areas Outside Leicester City
For shapes drawn outside Leicester City, building counts come from OpenStreetMap via the Overpass API. These don't use the UPRN database and show no data source icon.
| Region | Data Source | Icons Shown |
|---|---|---|
| Leicester City | OS Open UPRN (DuckDB) | and/or |
| All other locations | OpenStreetMap / Overpass | No icon |
Where You See These Icons
| Location | What appears |
|---|---|
| Dashboard → Buildings column | or next to the count |
| Dashboard footer → Data Sources Legend | Hover to see a tooltip explaining both icons |
Common Questions
Why do some streets have no icon?
They're outside Leicester City and use OpenStreetMap data, which doesn't go through the UPRN database at all.
Can a single street show both ⚡ and 🔍?
Yes. If the same street appears in multiple shapes — one fully covering it and another partially — both icons can appear in the combined row.
Is one method more accurate than the other?
Both are equally accurate for their purpose. Fast Path uses a pre-verified total — it's not an estimate, it's the real count. Precision Scan counts addresses individually, which is equally exact. The only difference is speed.
Does any of this data leave my browser?
No. Both the Street Summary Index and Individual Address File are loaded into DuckDB-Wasm, which runs entirely in your browser. No address data is ever sent to a server.
Can I choose which method to use?
No — and you don't need to. AGMT automatically picks the fastest accurate path for each street. It's always optimal.
For more on the data behind AGMT, see How It Works and Data Attribution & Licensing.