WaveFormDocs
Getting started

Overview

What WaveForm is, what each product does, and which one you probably want.

WaveForm is an infrastructure intelligence platform. We turn imagery and 3D scans — across satellite, drone, and mobile-LiDAR scales — into queryable, versioned, georeferenced site intelligence. Robots consume the result through a signed delivery layer called the nav-pack registry.

Three products, one timeline

WaveForm is organized around a cascade pipeline: data flows from the cheapest, broadest sensor (satellite) toward the most precise (mobile LiDAR), with each tier triggering work in the next.

Tier 1

Cloud platform

Satellite change-detection, drone inspection, scene rendering, reconstruction queue.

Tier 2

WaveForm Field (mobile)

iOS app for boots-on-the-ground capture with iPhone LiDAR. Offline-first, queue-and-sync.

Tier 3

Nav-pack registry

Signed, versioned navigation packs delivered to autonomous robots in the field.

Pick your path

WaveForm serves three roles. Each has a different starting page:

  • Surveyor / inspector. Install WaveForm Field, walk a site, upload bundles. Start with Install + sign in.
  • Robotics integrator. Register a fleet, fetch nav- packs, render them on-board. Start with Connecting a robot.
  • Platform admin / developer. Build automations on top of the HTTP API. Start with API reference.
If in doubt, read the quickstart
The quickstart walks the whole pipeline end-to-end in under 30 minutes. Most of what each role does fits inside it as a single step — easier to understand the system once you’ve seen it run.

How the pieces fit together

At the top, satellites monitor sites continuously and emit change alerts. A change alert can dispatch a drone (or schedule a human capture). The drone/mobile capture lands on the platform, gets reconstructed into a 3D scene with metrology-grade dimensions, and becomes a snapshot in time for that site.

Snapshots are the input to nav-packs. A nav-pack is the machine-readable contract a robot needs to operate at a site: it carries layers (map, lanelet, semantic, signed manifest), an autonomy tier (visual-only through full L4), and a cryptographic signature. The robot pulls the latest pack for its area + profile and uses it offline.

Auth model
Human users authenticate with Cognito (email + magic link). Robots authenticate with device tokens (HS256 JWTs) — see Device tokens. Both paths produce the same customer_id scoping downstream, so ACL is uniform.

What to do next

Docs — WaveForm — WaveForm