How the rating works
A screening or a survey
| Screening Condition Rating | ASTM D6433 PCI (field survey) | |
|---|---|---|
| What it is | A desktop rating from aerial imagery | A walked survey of sample units by the ASTM D6433 method |
| Scale | SCR 1 (very poor) to SCR 5 (very good) | PCI 0 (failed) to 100 (good) |
| Evidence | Every year of open imagery over the lot, software readings and an analyst’s marks | Distresses measured on site by type and severity, with photographs |
| Who decides | The software proposes; an Avok analyst sets the rating | Trained inspectors record; an engineer signs off |
| Use it for | Ranking a portfolio and choosing where to survey | Repair design, tendering and capital plans for one lot |
The five ratings
- SCR 5 – Very good
Surface looks uniform and dark. No potholes, patches, alligator areas or ponding are visible. Striping is clear. Often a recent paving or sealing.
- SCR 4 – Good
Light surface ageing or a few small patches. No potholes or alligator areas are visible. Striping is mostly clear.
- SCR 3 – Fair
Visible ageing over much of the surface, several patches, or isolated alligator areas or ponding. Striping is faded in places.
- SCR 2 – Poor
Alligator areas or patching over a large part of the surface, potholes visible, or ponding in several places. Striping is faded or missing over much of the lot.
- SCR 1 – Very poor
Alligator areas across most of the surface, several potholes or failed patches, and extensive ponding or debris. Striping is largely missing.
Survey priority
Each rated lot also gets a survey priority. Potholes raise it one step because they are a safety concern.
- High
- Schedule a field or drone survey within 30 days.
- Medium
- Schedule a field or drone survey within six months.
- Low
- Survey in the next annual cycle, or re-screen when new imagery is published.
How the software reads the imagery
The paved outline, less exclusions, is laid over the open imagery of every year that covers the site: Ontario’s provincial imagery (16 or 20 cm per pixel) and, in the United States, state programmes (7.5 to 30 cm; imagery finer than 12 cm is averaged to about 15 cm before it is read) and NAIP (30 cm to 1 m). Planting is found from the near-infrared band (or from colour where there is none), and cars, shadows and other objects from their size and contrast with the asphalt; they are left out, with a margin, together with painted lines and an 80 cm strip along kerbs.
On the pavement left visible, thin dark lines at least 1 m long are traced and measured. Ruler-straight lines that all point the same way but not along the parking layout are the shadows of light poles and are left out.
The line density (metres of line per 100 m² of visible pavement) of the newest year gives the proposed rating: bands at 3, 6.5, 15 and 22 m per 100 m². A weathered surface is not rated 5, and a density that grew markedly since the comparison year (the newest at least three years older) lowers the rating one step, unless the older imagery is coarser or washed out.
Cracks do not close by themselves: when the newest year shows markedly fewer lines and the surface is not darker (no sign of resurfacing), the comparison year sets the rating. Imagery where at least 15% of the visible pavement is at full brightness is washed out, and lines there are undercounted. Either lowers the confidence.
An Avok analyst reviews the readings with the imagery and confirms or changes the rating.
What a screening cannot see
- Cracks are traced only on imagery of 25 cm per pixel or finer. NAIP (United States, 30 cm to 1 m) gives the share visible, the tone and the markings, not cracks.
- Cracks narrower than about one pixel (15 to 20 cm as read) are only seen when they show as a dark line; hairline cracks, ravelling and rutting are not measured. Coarser or washed-out imagery shows fewer cracks.
- A darker surface is read as resurfacing only between flights of the same source, only if the cracks largely went, and not when many lines (15 m per 100 m² or more) show within three years after it: that is more likely a seal coat or a difference between flights. Tones from different sources (seasons, cameras) are not compared.
- Pavement under cars, trees and shadows is not read; the share visible is stated.
- Sealed cracks, paving joints and tyre marks read as lines; light-pole shadows that run along the layout are counted as lines.
- The rating bands were set on five lots in York Region in October 2026 and will be checked against field surveys.
How accurate is it?
The rating bands were set in October 2026 on five parking lots in York Region, Ontario. They will be checked against walked surveys on at least ten lots, and the agreement will be published on this page. Until then, the software’s rating is a proposal that an analyst confirms or changes.
The imagery
Ontario: Geospatial Ontario’s imagery, flown region by region about every five years since 2013, at 16 or 20 cm per pixel. United States: USDA NAIP in every state (30 cm to 1 m) and state programmes at 7.5 to 30 cm where their terms allow commercial use. Imagery finer than 12 cm is averaged to about 15 cm before it is read, the pixel size the rating bands were set on.