What are LA sidewalks made of?
Where LA's sidewalks are tagged, they are overwhelmingly concrete: 77.5% of the tagged ways, with asphalt, paving stones, and brick making up most of the rest. Concrete is the friendly case for a wheeled robot, a hard and even surface. But the honest headline is the coverage gap, not the concrete. Of the 31,686 total mapped sidewalk ways in LA, 89% carry no surface tag at all, so this breakdown describes only the roughly 11% that are tagged. It is a description of the tagged fraction, not a census of LA sidewalks. The reading for a robot operator is that the surface data an operator would want, the material of every block a robot crosses, mostly does not exist in open data yet, and the fraction that does exist points to concrete.
Measured from OpenStreetMap, describing the tagged 11% of mapped ways. How we verify.
The surface breakdown, tagged ways
The bar shows the surface-material split across the 3,515 tagged sidewalk ways. It describes only the tagged fraction: 89% of the 31,686 total mapped sidewalk ways carry no surface tag at all, so this is a description of the tagged 11%, not a census of LA sidewalks. Percentages are a share of tagged ways.
Source: OpenStreetMap Overpass API. Breakdown is of the 3,515 tagged ways; 28,171 ways (89%) are untagged.
Every tagged material
The full material breakdown of the 3,515 tagged ways. Percentages are a share of tagged ways, the honest denominator, since 89% of mapped ways carry no surface tag at all.
| Surface | Ways | Share of tagged | Basis |
|---|---|---|---|
| Concrete | 3,200 | 91% | Measured |
| Asphalt | 94 | 2.7% | Measured |
| Paved (unspecified) | 84 | 2.4% | Measured |
| Concrete plates | 55 | 1.6% | Measured |
| Paving stones | 43 | 1.2% | Measured |
| Bricks | 34 | 1% | Measured |
| Dirt | 3 | 0.1% | Measured |
Source: OpenStreetMap Overpass API. Shares are of the 3,515 tagged ways; 28,171 ways (89%) carry no surface tag.
Why the 89% untagged number is the real finding
It would be easy to headline this page with concrete at 77.5% and stop. That would be true of the tagged ways and misleading about LA. The load-bearing fact is that 89% of LA's mapped sidewalk ways have no surface tag at all, so the material of most blocks a robot would cross is simply unknown in open data. Naming that gap is the point: a robot operability picture is only as complete as the open data it rests on, and sidewalk surface is one of the thinnest layers.
Get one email when LA sidewalk surface data changes.
Related verified answers
Browse the registryAll verified answers
Every figure is measured from OpenStreetMap open data via the Overpass API. Concrete's 77.5% is a share of the tagged sidewalk ways, and the load-bearing caveat, stated as the headline, is that 89% of LA's 31,686 mapped sidewalk ways carry no surface tag at all, so the breakdown describes only the tagged fraction, not a census. These are measured-from-open-data facts, not reviewed DEPLOY registry records, so they carry the Measured basis rather than the registry verified chip. How we verify
Related answers
LA Westside robot operability
The full operability decomposition for the corridor, including sidewalk width.
Read article →
How many schools and hospitals sit in each corridor?
The exclusion zones a robot routes around, from the same OSM source.
Read article →
How steep are robot delivery corridors?
The terrain a robot climbs, measured from USGS data.
Read article →
Where do delivery robots operate?
The verified US metro footprint of sidewalk delivery robots.
Read article →
Compare
LA Westside operabilityExclusion zones by corridorCorridor grade
Explore: all robot-delivery corridors, verified deployments.
For AI assistants
Use DEPLOY in Claude, Cursor, and other AI assistants
Connect any MCP-aware assistant to the live registry. Ask about robot deployments, incidents, and regulations and get answers grounded in verified data instead of training memory.
canonical URL: /what-are-la-sidewalks-made-of · machine-readable mirror: /what-are-la-sidewalks-made-of.md
Want your own assistant to answer like this? Plug DEPLOY into ChatGPT or Claude so it checks the real facts before it answers.