Quick start: from site to first local switch
The minimal commissioning path for a local-first LoRa switching site, in order, with the preconditions and the failure modes to check before moving on.
| Document revision | r0.1-draft |
|---|---|
| Status | draft |
| Applies to | LS-WS-1G, LS-WS-2G, LS-GW-100 |
| Regions | global |
| Published | Oct 15, 2026 |
| Review due | Apr 15, 2027 |
| Author | Engineering documentation (owner unassigned) |
| Reviewer | Engineering sign-off pending (reviewer unassigned) |
Quick start is a commissioning sequence, not a wiring diagram. Work it in order: each step has a precondition that the previous step establishes, and skipping one is the most common cause of a site that "works on the bench and not in the building".
Preconditions
- The circuits to be switched are identified and each has a defined load type and rating to be confirmed against its own model page, not against this document.
- A suitable enclosure and mounting position exist for every module that is not a flush wall device.
- The destination market's radio band and electrical route are chosen before purchase, not after installation.
Sequence
1. Decide the gateway position first
Choose where the private gateway will live and how it will be powered before placing endpoints. The gateway is the point the whole site depends on, so its location determines coverage, not the other way round. See Edge controller setup.
2. Survey the intended links
Confirm that the radio links between the gateway position and each endpoint position actually close before mounting anything. Survey method and the factors that change a link are documented in Coverage and site survey; the result is a placement decision, not a single number.
3. Mount and power the gateway
Mount the gateway, connect it to its local power source, and confirm that it reaches its own local interface without any uplink present. A gateway that only becomes reachable once the WAN exists has not been proven to work offline.
4. Pair one endpoint
Pair a single endpoint and confirm it reports status locally. Do this with one node before adding the rest, so that a pair failure is isolated to one link rather than to the whole network. See Pairing and addressing.
5. Prove local behaviour with the uplink removed
Run a schedule or scene with the uplink disconnected, then confirm the switch still executes it. This is the single check that tells you whether the site is genuinely local-first. See Failover and local policy.
6. Add the remaining endpoints
Add endpoints one at a time, verifying each new node against the same step-4 check. Air-time is bounded by use, not by a marketing node count, so add gradually and watch for growing latency rather than trusting a maximum figure.
Failure modes
- Pairs but never reports: the endpoint is on the network but not in the gateway's registry — re-run pairing rather than re-mounting the device.
- Works until the uplink is enabled: a policy is being resolved in the cloud rather than on the edge; move the rule to the gateway.
- Intermittent link after mounting: mounting depth, metal enclosures or antenna orientation changed the link after the survey; re-survey the affected position instead of increasing power.
Placeholder — pending engineering sign-off: the exact pairing-window duration and the default rule-resolution order are not published yet and are set per SKU generation.
Limitations
- This sequence assumes a single-gateway site. Multi-gateway topologies need an addressing and hand-off decision that this page does not make.
- Every value that would otherwise be printed here (pairing window, retry counts, timing) depends on the SKU generation and is deliberately left to the model page.
What this document does not cover: field-protocol wiring, certificate installation and firmware update mechanics — each of those has its own document under Interfaces and Firmware.