Every good version of this kind of report is built on one real structural asymmetry inside the product. Here is LiDAR's.
Firing a laser pulse, catching its reflection with a photodiode, and converting the elapsed time into a distance is decades-old, well-understood physics — the core technique dates to 1960s-era airborne surveying instruments. At short range, in good light, with a cooperative (reasonably reflective) target, building a device that does this is genuinely accessible: it is why dozens of companies, the great majority of them Chinese, can and do sell mechanical-spinning or simple solid-state LiDAR modules for robot vacuum cleaners, warehouse AMRs and entry-level mapping today, at unit prices that have fallen by an order of magnitude over the past decade. Failure here is visible and immediate — the unit either returns a coherent point cloud on the bench or it does not.
The same physics, pushed to 200+ metres, through rain or fog, in direct sunlight, against a low- reflectivity target like dark clothing, is an entirely different problem: the single photon your own laser reflects back must be distinguished from an overwhelming background of ambient solar photons hitting the same detector — a signal-to-noise problem — while the laser's peak power is hard-capped by the eye-safety limit in §1, and the whole assembly must survive fifteen years of vibration, temperature cycling and vibration at an automotive-grade reliability bar, then be certified to fail safely under ISO 26262 when it eventually does fail. Both layers share the same failure signature: invisible and delayed. A marginal detection-margin design does not fail on a datasheet; it fails as a missed detection in the one rare combination of range, weather and target reflectivity that the bench test never happened to recreate — exactly when it is most expensive to discover.
| The easy half — the point cloud | The hard half — the detection margin | |
|---|---|---|
| What it is | Pulse-and-catch ranging at short range, good conditions, cooperative targets | Extracting a real signal from solar noise at long range, in weather, against low-reflectivity targets — inside a fixed eye-safety power budget |
| Certification gate | A working bench demo | ISO 26262 ASIL-B/D functional safety, IATF 16949 manufacturing quality, and OEM-specific design validation — each repeated per vehicle platform |
| How it fails | Visible immediately — no return signal, no point cloud | Invisible and delayed — a missed detection that only shows up in the rare real-world combination of range, weather and target reflectivity a bench test does not recreate |
| Who can attempt it | Any competent optics/electronics shop — dozens of Chinese suppliers now do, at commodity prices | A much smaller population with real automotive functional-safety and manufacturing-quality engineering depth, proven at six- and seven-figure unit volumes |
| Where the margin sits | Thin and falling — see §9's consolidation record | Real, but concentrated in whichever company has actually converted engineering depth into automotive design wins at scale — see §9 |
Making light bounce back is the easy part. Finding the one photon that matters, inside a fixed eye-safety power ceiling, at automotive cost and automotive reliability, is the product. Everything else in this report — the value ladder in §5, the consolidation record in §9, the ratings in §11 — is really a question of how much of a given company's revenue sits on the detection-margin side of this line, versus the commodity-point-cloud side.
